CS Disco Inc (LAW) 2023 Q1 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to CS Disco's First Quarter of Fiscal Year 2023 Conference Call. (Operator Instructions) I would now like to hand the conference over to your first speaker today, Aleksey Lakchakov, Investor Relations. Please go ahead.

女士們、先生們，感謝您的耐心等待，歡迎參加CS Disco 2023財年第一季電話會議。 （操作說明）現在我將會議交給今天的第一位發言人，投資者關係部的Aleksey Lakchakov先生。請開始發言。

Good afternoon and thank you for joining us on today's conference call to discuss the financial results for DISCO's first quarter of 2023. With me on today's call are Kiwi Camara, DISCO's Co-Founder and Chief Executive Officer; and Michael Lafair, DISCO's Chief Financial Officer.

下午好，感謝各位參加今天的電話會議，共同探討DISCO公司2023年第一季的財務表現。今天與我一同參加會議的還有DISCO公司的共同創辦人兼執行長Kiwi Camara先生，以及DISCO公司的財務長Michael Lafair先生。

Today's call will include forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995, including, but not limited to, statements regarding our financial outlook and future performance or future capital expenditures, market opportunity, market position, product strategy and growth opportunities and developments in the legal technology industry. In addition to our prepared remarks, our earnings press release, SEC filings and a replay of today's call can be found on our Investor Relations website at ir.csdisco.com.

今天的電話會議將包含符合1995年《私人證券訴訟改革法案》安全港條款定義的前瞻性陳述，包括但不限於有關我們財務展望和未來業績或未來資本支出、市場機會、市場地位、產品戰略以及法律科技行業增長機會和發展的陳述。除我們準備好的演講稿外，您還可以在我們的投資者關係網站ir.csdisco.com上找到我們的盈利新聞稿、提交給美國證券交易委員會（SEC）的文件以及今天電話會議的錄音回放。

Forward-looking statements involve known and unknown risks and uncertainties that may cause our actual results, performance or achievements to be materially different from those expressed or implied by the forward-looking statements. Forward-looking statements represent our management's beliefs and assumptions only as of the date made. Information on factors that could affect the company's financial results is included in its filings with the SEC from time-to-time, including the section titled Risk Factors in the company's quarterly report on Form 10-Q for the quarter ended March 31, 2023, filed with the SEC on May 10, 2023, and the company's annual report on Form 10-K for the year ended December 31, 2022, filed with the SEC on February 24, 2023.

前瞻性陳述涉及已知和未知的風險和不確定性，可能導致我們的實際績效、表現或成就與前瞻性陳述中明示或暗示的內容有重大差異。前瞻性陳述僅代表我們管理階層截至陳述之日的信念和假設。可能影響公司財務表現的因素資訊包含在本公司不時向美國證券交易委員會（SEC）提交的文件中，包括本公司截至2023年3月31日止季度的10-Q季度報告（於2023年5月10日向SEC提交）中題為「風險因素」的部分，以及本公司截至2022年5月10日向SEC提交）中題為「風險因素」的部分，以及本公司截至2022年12022年

In addition, during today's call, we will discuss non-GAAP financial measures. These non-GAAP financial measures are in addition to and that a substitute for or superior to, measures of financial performance prepared in accordance with GAAP. Reconciliation between GAAP and non-GAAP financial measures and a discussion of the limitations of using non-GAAP measures versus our closest GAAP equivalents is available in our earnings release.

此外，在今天的電話會議中，我們將討論非公認會計準則（非GAAP）財務指標。這些非GAAP財務指標是依照公認會計準則（GAAP）編製的財務績效指標的補充，並非替代或優於後者。 GAAP和非GAAP財務指標之間的調節表以及使用非GAAP指標相對於最接近的GAAP指標的局限性討論，請參閱我們的盈利報告。

And with that, I'd like to turn the call over to Kiwi.

接下來，我想把電話交給 Kiwi。

Thanks, Aleksey. Good afternoon, everyone, and welcome to our earnings call for the first quarter of fiscal year 2023. Revenue for Q1 2023 was $33.1 million. Adjusted EBITDA was negative $13 million and total customer count grew to 1,388. Revenue was above the high end of our guidance range and adjusted EBITDA significantly exceeded our guidance range. We have made steady progress on the operational changes that we expect will lead to profitability, including the launch of our new office in Delhi.

謝謝，阿列克謝。大家下午好，歡迎參加我們2023財年第一季財報電話會議。 2023財年第一季營收為3,310萬美元。調整後EBITDA為負1,300萬美元，客戶總數成長至1,388家。營收高於我們預期範圍的上限，調整後EBITDA也顯著超出預期範圍。我們在營運改革方面取得了穩步進展，預計這些改革將有助於實現盈利，其中包括我們在德里開設的新辦事處。

As Michael will share in more detail, this has allowed us to improve our adjusted EBITDA outlook for this year and pull forward our anticipated time line for achieving positive adjusted EBITDA from Q4 2024 to Q3 2024. Q1 was a very exciting quarter for us for both the product and the sales and marketing perspective. First, let's talk about product. We are huge believers in AI at DISCO. Recent advances in large language models that you will have read about in the news as well as less heralded but equally important advances in technologies like Vector search, represent, in my opinion, the most important technology development since the mass adoption of the Internet in the 90s.

正如邁克爾將要詳細闡述的那樣，這使我們能夠提高今年的調整後 EBITDA 預期，並將實現調整後 EBITDA 為正的預期時間從 2024 年第四季度提前至 2024 年第三季度。第一季對我們來說是一個非常令人振奮的季度，無論從產品還是銷售和行銷的角度來看都是如此。首先，我們來談談產品。 DISCO 堅信人工智慧的潛力。您可能已經在新聞中讀到過，大型語言模型的最新進展，以及向量搜尋等鮮為人知但同樣重要的技術進步，在我看來，代表了自 90 年代互聯網普及以來最重要的技術發展。

These technologies, which AI labs throughout the world have been working on for years are finally ready to be incorporated into products that can deliver real value to everyday users. In Q1, we introduced Cecilia, on integrated AI chatbot for large-scale e-discovery. Cecilia will allow lawyers to ask natural language questions and receive narrative answers based on information in the documents in a customer's private DISCO Ediscovery database. To support her answers, Cecelia will provide citations to documents in the database, allowing lawyers to verify her answers, site evidence in support of them and use the documents as a starting point for further investigation.

全球人工智慧實驗室多年來致力於研發的這些技術，終於可以應用於產品中，為日常用戶帶來真正的價值。第一季度，我們推出了 Cecilia，一款用於大規模電子取證的整合式人工智慧聊天機器人。 Cecilia 允許律師使用自然語言提問，並根據客戶私人 DISCO 電子取證資料庫中的文件資訊獲得敘述性答案。為了佐證其答案，Cecilia 會提供資料庫中文件的引用，方便律師驗證答案、尋找證據，並將這些文件作為進一步調查的起點。

Cecilia is differentiated by her ability to answer questions based on private data rather than the public Internet or public corpora of books and articles. To do so at Ediscovery scale in databases that may involve millions or tens of millions of documents, to cite evidence in support of her answers and to do all this while integrated into a secure platform that lets lawyers do other legal work, like ingesting data, exploring and analyzing documents, orchestrating large-scale reviews and preparing productions or disclosures.

Cecilia 的獨特之處在於，她能夠基於私人資料而非公開網路或公開書籍文章庫來回答問題。她能夠在包含數百萬甚至數千萬份文件的資料庫中進行大規模的電子取證，並引用證據來支持她的答案。所有這些都整合在一個安全平台上，律師可以在這個平台上進行其他法律工作，例如資料導入、文件探索和分析、組織大規模審查以及準備證據揭露文件。

We launched Cecelia at Legal Week in New York and allowed customers and prospects to begin testing her asking her questions themselves. At these initial demonstrations and in conversations with our top customers, we are hearing fantastic feedback. Lawyers immediately see the power of Cecilia to accelerate the process of understanding the facts and finding evidence in legal disputes and investigations. I would have loved to have had Cecilia help me on cases when I was a practicing attorney. We anticipate that Cecilia will be generally available in 2023 after an initial staged rollout.

我們在紐約法律週期間發布了Cecelia，並讓客戶和潛在客戶親自體驗並向她提問。在這些初步演示以及與頂級客戶的交流中，我們收到了非常棒的回饋。律師們立刻感受到Cecelia的強大功能，它能加速理解事實，並在法律糾紛和調查中尋找證據。如果我當初還是執業律師時，Cecelia能幫我處理案件，那該有多好！我們預計，經過初期的分階段推廣後，Cecelia將於2023年正式上線。

We will provide further information on pricing at that time as well. Cecilia is both the first step in our productization of modern AI technologies and a continuation of DISCO's investment in AI that dates back almost to our founding. I have often asked about DISCO AI and the direction we are going. What is DISCO AI exactly? What does it do? How is it differentiated? DISCO AI is a collection of technologies that is fully embedded in our platform, and that is productized to allow lawyers to seamlessly use AI in the course of legal work. DISCO AI powers Predicted Tags, which helps lawyers identify documents related to particular legal issues.

屆時我們將提供更多關於定價的資訊。 Cecilia 既是我們現代人工智慧技術產品化的第一步，也是 DISCO 自成立以來對人工智慧持續投入的延續。我常被問到 DISCO AI 以及我們的發展方向。 DISCO AI 究竟是什麼？它的功能是什麼？它的獨特之處是什麼？ DISCO AI 是一系列完全嵌入我們平台的技術集合，其產品化旨在讓律師在法律工作中無縫使用人工智慧。 DISCO AI 為預測標籤提供支持，幫助律師識別與特定法律問題相關的文件。

Our Predicted Tags technology predicts how a lawyer would tag a document based on how lawyers have tagged other documents in the past. For every document, Predicted Tags generates scores for each tag that indicate how likely DISCO AI believes the document is to be tagged a certain way. Users can then search by these scores to find documents that are likely to be what they're looking for. They can incorporate Predicted Tag scores as part of more complicated searches in filters and search and visualization. Predicted tags can help them review documents more quickly by telling them what they should be looking for in a document.

我們的預測標籤技術能夠根據律師以往的文檔標記方式，預測律師會如何標記目前文檔。對於每份文檔，預測標籤都會為每個標籤產生分數，這些分數表明 DISCO AI 認為該文檔被標記的可能性有多大。使用者可以根據這些分數進行搜索，找到可能符合他們需求的文件。他們還可以將預測標籤分數整合到更複雜的搜尋中，例如在篩選器、搜尋和視覺化功能中。預測標籤能夠幫助使用者更快審閱文檔，因為它能告訴使用者應該在文檔中尋找哪些內容。

Predicted tags can be used to sort documents for review, ensuring that lawyers are looking at more of the important documents per unit time invested and that lawyers find more of the important documents more quickly. And Predicted Tags can be used to do real-time quality control on human reviews, identifying areas where human reviewers disagree with DISCO AI's recommendation. DISCO AI powers topic clustering with automatic indexing, which can take millions of documents and organize them by relevant topics labeled using terms and phrases used in the actual documents.

預測標籤可用於對文檔進行排序以供審閱，確保律師在單位時間內審閱更多重要文檔，並更快找到更多重要文檔。此外，預測標籤還可用於對人工審閱進行即時品質控制，識別人工審閱者與 DISCO AI 建議存在分歧之處。 DISCO AI 利用自動索引功能支援主題聚類，能夠處理數百萬份文檔，並根據文件中使用的術語和短語，按相關主題對其進行分類。

This allows lawyers to understand what topics are covered by documents in a database before beginning a review. It helps lawyers learn the particular language or jargon used in a database. And in conjunction with filtering and search visualization, it lets lawyers see what topics are covered in particular subsets of a database, for example, in the documents from a particular witness or party. And DISCO AI powers cross matter AI, which allows lawyers to use models trained on earlier matters on new matters.

這使得律師能夠在開始審查之前了解資料庫中文件涵蓋的主題。它有助於律師學習資料庫中使用的特定語言或術語。結合篩選和搜尋視覺化功能，律師可以查看資料庫特定子集涵蓋的主題，例如，特定證人或當事人的文件。此外，DISCO AI 還支援跨案件 AI，使律師能夠將基於先前案件訓練的模型應用於新案件。

This allows lawyers and ultimately, their clients to get leverage on the work they invested in prior matters to accelerate review on new matters that involve similar facts or similar legal issues. Just as a lawyer who has handled a certain kind of case many times in the past, can be more efficient than a lawyer who is handling that kind of case for the first time, DISCO AI can become more and more effective as it learns across matters, not just within a single matter. When we think about DISCO's product road map, we think about modern AI technologies, unlocking features in 4 areas.

這使得律師及其客戶能夠利用先前在類似案件中投入的時間和精力，加快對涉及類似事實或法律問題的新案件的審查。正如一位處理過多種類型案件的律師比初次處理此類案件的律師效率更高一樣，DISCO AI 也能隨著在不同案件中的學習而變得越來越有效，而不僅限於單一案件。在規劃 DISCO 的產品路線圖時，我們著眼於現代人工智慧技術，並致力於在四個領域解鎖新功能。

First is question-answering of which Cecilia is an example. Second is document analysis. We think modern AI can be used to tag and parse documents, labeling, summarizing and extracting data as instructed in natural language on a document-by-document basis. Third is drafting and document generation and fourth is providing a natural language user interface for DISCO's platform, where a user can simply ask the computer to do what they want, making our products even easier to use.

首先是問答功能，Cecilia 就是一個例子。其次是文檔分析。我們認為現代人工智慧可以用於標記和解析文檔，根據自然語言指令逐個文檔地進行標註、摘要和資料擷取。第三是文檔起草和生成。第四是為 DISCO 平台提供自然語言使用者介面，使用者只需發出指令即可讓電腦執行所需操作，使我們的產品更加易於使用。

Our ability to take modern AI and integrate it into a broad platform where lawyers already do their work and through which we have access to private data and work product to power our AI models, represents a material advantage for DISCO relative to pure-play AI companies. To be clear, these are ideas for future product development, not product features that we have today. But we believe this framework can help you understand how we are thinking about productizing modern AI technologies into our platform and give you insight into why we think these technologies will be impactful in accelerating and automating legal work and improving the experience of practicing law for our customers.

我們能夠將現代人工智慧技術整合到一個律師們日常工作的廣泛平台中，並透過該平台獲取私有資料和工作成果來驅動我們的人工智慧模型，這為DISCO相對於純粹的人工智慧公司帶來了實質的優勢。需要明確的是，這些是未來產品開發的構想，而非我們目前已有的產品功能。但我們相信，這個框架能夠幫助您理解我們如何將現代人工智慧技術產品化並融入我們的平台，並讓您深入了解我們為何認為這些技術能夠顯著加速和自動化法律工作，並改善客戶的執業體驗。

Artificial intelligence is not the only element of our platform we are excited to discuss. This quarter, we also announced the release of time lines in Case Builder. Time lines allows lawyers to collaboratively create events and facts, link them to depositions, documents and other evidence in the DISCO platform and elsewhere and assemble them into time lines and chronologies that the legal team uses to understand, develop and present their case. Time lines is part of our strategy to be the end-to-end legal technology provider. Whereas Ediscovery generally occurs well into a legal matter, Case Builder is now well positioned to capture the matter in the moment a client comes to a lawyer with an issue.

人工智慧並非我們平台唯一值得探討的亮點。本季度，我們也宣佈在 Case Builder 中推出時間軸功能。時間軸功能允許律師協作創建事件和事實，並將它們與 DISCO 平台及其他平台上的證詞、文件和其他證據關聯起來，最終形成時間軸和年表，供法律團隊理解、發展和陳述案情。時間軸功能是我們成為端到端法律技術提供者策略的一部分。以往，電子取證通常在案件審理後期才會進行，而 Case Builder 現在能夠在客戶就案情諮詢律師的那一刻就迅速掌握案件資訊。

Our customers have been asking for the capability that time lines provides for some time, and we are very pleased to bring it to market. There are other under-the-hood features we released this quarter as well that are core to improving DISCO's capabilities. One example of this is live reindexing. Periodically, as we develop our products, we make changes to the underlying data or search architecture that enable improved performance or new capabilities. Live reindexing automates the process of moving databases that are on older versions of our platform to the latest versions.

我們的客戶一直以來都希望我們能提供時間軸功能，我們很高興現在能將其推向市場。本季我們也發布了其他一些底層功能，這些功能對於提升 DISCO 的整體效能至關重要。例如，即時重新索引就是其中之一。隨著我們產品的不斷開發，我們會定期對底層資料或搜尋架構進行更改，以提升效能或新增功能。即時重新索引功能可以自動將執行在舊版平台上的資料庫遷移到最新版本。

Previously, this was a time-consuming task managed by human engineers on a case-by-case basis. We believe live reindexing will improve the user experience for our clients, especially those clients whose matters tend to last for multiple years on our platform. On the sales and marketing side, Q1 was a productive quarter. The highlight of the quarter was strong funnel activity on an aggregate and per rep basis. Across our measures such as meetings, near-term opportunities and wins, we have been seeing acceleration since the start of the year. This is telling us that the operational initiatives we have been implementing internally and that we discussed on the last earnings call are starting to drive results.

此前，這項工作耗時費力，需要工程師逐一案例處理。我們相信，即時重新索引將改善客戶的用戶體驗，尤其是那些在我們平台上處理案件長達數年的客戶。在銷售和行銷方面，第一季成果豐碩。本季的亮點是整體和每位銷售代表的銷售漏斗活動都非常活躍。從會議數量、近期商機和成交量等指標來看，自年初以來，我們看到各項指標都在加速成長。這表明，我們一直在內部實施並在上次財報電話會議上討論的營運措施已開始產生成效。

In our industry, neither we nor our customers know how large a matter will ultimately be, but we do control our ability to go after every deal. Our improving pipeline metrics demonstrate success in that effort. While we are pleased with the continued improvement in pipeline metrics and the resulting growth in total customers, we continue to see less usage of our platform on very large matters across all our products. Additionally, we continue to see pressure from our largest customers to reduce their spend on our platform by reducing the amount of data in our platform, more aggressively utilizing lower-cost DISCO offerings like DISCO ECA and renegotiating total spend.

在我們的行業中，無論是我們還是客戶，都無法預測最終案件的規模，但我們有能力爭取每一筆交易。我們不斷提升的銷售管道指標證明了我們在這方面的努力取得了成功。儘管我們對銷售通路指標的持續改善以及由此帶來的客戶總數成長感到欣慰，但我們仍然發現，在所有產品中，我們平台在處理超大型案件方面的使用率有所下降。此外，我們持續感受到來自最大客戶的壓力，他們要求我們減少在平台上的支出，具體措施包括減少平台上的資料量、更積極地利用成本更低的DISCO產品（例如DISCO ECA）以及重新協商總支出。

Given our usage-based business model, this pressure on usage levels can mean that revenue does not grow in line with the improvement we are seeing in the pipeline metrics that we more directly control. We can sell lots of new customers and add lots of new matters without necessarily seeing the net increase in usage that would drive revenue growth. We continue to believe that the right strategy is to continue adding new customers to our platform, drive increased adoption of our platform across more matters at each customer and drive customers to adopt more DISCO products.

鑑於我們採用的是基於使用量的商業模式，使用量方面的壓力可能導致收入成長無法與我們更直接可控的業務指標的改善同步。我們可以新增大量客戶，增加大量新案件，但未必能獲得足以推動營收成長的淨使用量成長。我們仍然認為，正確的策略是繼續為我們的平台增加新客戶，推動每位客戶在更多案件中使用我們的平台，並引導客戶採用更多DISCO產品。

Over time, we believe this will result in increased usage and consequently increased revenue. In Q1, we launched our law better marketing campaign. Law better announced to the world that DISCO is here to make the law work better for everyone. We launched on digital streaming services with full-length commercials as well as the targeted digital blips on social media. You might have even seen our Lady Justice or Lady J on your own TV. If you have not, you should look up (inaudible). Our customers and new prospects have been very positive about this campaign. We are seeing high engagement with our content across all media channels.

我們相信，隨著時間的推移，這將帶來更高的使用率，進而增加收入。第一季度，我們推出了「讓法律更有效率」行銷活動。 「讓法律更有效率」向世界宣告，DISCO 致力於讓法律服務更好地惠及所有人。我們透過數位串流服務平台發布了完整版廣告，並在社群媒體上投放了精準的數位廣告。您甚至可能在電視上看到我們的“正義女神”或“J 女士”。如果您還沒看到，不妨上網搜尋一下（聽不清楚）。我們的客戶和潛在客戶對此次行銷活動反應熱烈。我們看到，在所有媒體管道上，我們的內容都獲得了很高的互動率。

On our website, visitors are spending 15x more time on our campaign pages versus site-wide averages. We are seeing engagement with some of the largest legal buyers in the world coming to our website, exploring our products and reaching out for conversations. Both Cecilia and Law better were on full display at Legal Week in New York this past March, where we saw strong attendance, engagement with our team and lead generation. One lawyer at the show told us that when he saw the Law better campaign on TV he brought his entire family in to watch.

在我們的網站上，訪客在活動頁面上的停留時間是網站平均水平的15倍。我們看到一些全球最大的法律產品採購商造訪我們的網站，瀏覽我們的產品並與我們聯繫洽談。今年三月，Cecilia和Law Better在紐約法律週上全面亮相，展會現場人氣爆棚，我們的團隊也積極參與互動，並獲得了大量潛在客戶。一位律師告訴我們，他在電視上看到Law Better的廣告後，就帶著全家一起來看了。

It's not often lawyers receive marketing attention. We are pleased with the excitement our campaign has generated among our customers and prospects. Finally, operationally, we are continuing our push to globalize with the opening of our new office in Delhi. Through our India presence, we will be able to provide 24/7 customer service and support to our customers worldwide and continue to scale out other business functions, including engineering. This team will augment our existing team of Discovians in North America and Europe and will be a tailwind on our path to profitability.

律師業很少會受到行銷的關注。我們很高興此次行銷活動在客戶和潛在客戶中引發瞭如此熱烈的反響。此外，在營運方面，我們正隨著德里新辦事處的成立，繼續推動全球化策略。透過我們在印度的業務佈局，我們將能夠為全球客戶提供全天候的客戶服務和支持，並持續拓展其他業務職能，包括工程部門。這支新團隊將壯大我們在北美和歐洲的現有團隊，並協助我們獲利。

With that, I will turn it over to Michael.

接下來，我將把麥克風交給麥可。

Thank you, Kiwi. In Q1 2023, revenue was $33.1 million, a decline of 4% year-over-year. The year-over-year decline in the quarter is attributable to the smaller matters we saw in the Review business. In discussing the remainder of the income statement, please note that unless otherwise specified, all references to our expenses, operating results and share count are on a non-GAAP basis. Our gross margin in Q1 was 76%, up from 74% in Q1 of the prior year. As we mentioned before, our gross margins fluctuate from period-to-period based on the nature of our customers' usage, for example, the amount and types of data ingested and managed on our platform.

謝謝您，Kiwi。 2023年第一季度，我們的營收為3,310萬美元，年減4%。本季營收年減主要歸因於評論業務中一些較小的項目。在討論損益表的其餘部分時，請注意，除非另有說明，所有費用、經營業績和股份數量的引用均基於非GAAP準則。第一季的毛利率為76%，高於去年同期的74%。正如我們之前提到的，我們的毛利率會根據客戶的使用情況而波動，例如，客戶在我們平台上接收和管理的資料量和資料類型。

We expect gross margin to continue to be within the bands we've historically seen. Sales and marketing expense for Q1 was $17.5 million or 53% of revenue compared to 45% of revenue in Q1 of the prior year. This represents an increase of approximately $1.9 million in the quarter year-on-year. The increase was driven by our head count investments in 2022 as well as our marketing campaign launched in Q1. Research and development expense for Q1 was $13.1 million or 39% of revenue compared to 32% of revenue in Q1 of the prior year.

我們預計毛利率將繼續保持在歷史水平範圍內。第一季銷售和行銷費用為1,750萬美元，佔營收的53%，去年同期為45%。這比去年同期增加了約190萬美元。成長主要源自於我們在2022年的人員增加以及第一季啟動的行銷活動。第一季研發費用為1,310萬美元，佔營收的39%，去年同期為32%。

This represents an increase of approximately $2.2 million in the quarter year-on-year. This growth was driven primarily by our engineering head count investments in 2022. General and administrative expense in Q1 was $8.6 million or 26% of revenue compared to 22% of revenue in Q1 of the prior year. This represents an increase of over $0.8 million in the quarter year-on-year. Operating loss in Q1 was $14 million, representing a margin of negative 42% compared to negative 25% in Q1 of the prior year.

本季營收年增約220萬美元。這一增長主要得益於我們在2022年增加的工程人員數量。第一季一般及行政費用為860萬美元，佔營收的26%，去年同期為22%。這比去年同期增加了80多萬美元。第一季營業虧損為1,400萬美元，利潤率為-42%，去年同期為-25%。

Adjusted EBITDA was negative $13 million in Q1, a margin of negative 39% compared to a margin of negative 23% in Q1 of the prior year. Net loss in Q1 was $12.1 million or negative 37% of revenue compared to a net loss of $8.6 million or negative 25% of revenue in Q1 of the prior year. Net loss per share for Q1 was $0.20 per share compared to $0.15 per share in Q1 of the prior year. Turning to the balance sheet and cash flow statement; we ended Q1 with $187.6 million in cash and cash equivalents.

第一季調整後 EBITDA 為負 1,300 萬美元，利潤率為負 39%，去年同期為負 23%。第一季淨虧損為 1,210 萬美元，佔營收的 37%，而去年同期淨虧損為 860 萬美元，佔營收的 25%。第一季每股淨虧損為 0.20 美元，而去年同期為 0.15 美元。資產負債表及現金流量表顯示，第一季末公司持有現金及現金等價物 1.876 億美元。

Operating cash flow for Q1 was negative $14.8 million compared to negative $11.4 million in the prior year. On May 9, we took cost-cutting actions that impacted a number of employees across all departments. These changes are reflected in the new adjusted EBITDA guidance we are about to provide. Now turning to the outlook; for Q2 2023, we are providing revenue guidance in the range of $31 million to $33 million and adjusted EBITDA guidance in the range of negative $12 million to negative $10 million.

第一季經營現金流為負1,480萬美元，而上年同期為負1,140萬美元。 5月9日，我們採取了成本削減措施，影響了各部門的部分員工。這些變更已反映在我們即將發布的新的調整後EBITDA預期中。展望未來，我們預計2023年第二季營收將在3,100萬美元至3,300萬美元之間，調整後EBITDA將在負1,200萬美元至負1,000萬美元之間。

For fiscal year 2023, we are reiterating our prior revenue guidance of $135 million to $145 million and raising our adjusted EBITDA guidance from a range of negative $42 million to $38 million to a range of negative $40 million to negative $36 million. We are continuing to target Q4 2023 adjusted EBITDA margin in the negative teens. Additionally, we are pulling forward the anticipated time line of our path to profitability by one quarter. We had previously shared that we expect the adjusted EBITDA positive exiting 2024. We now expect to be adjusted EBITDA positive in Q3 of 2024.

對於2023財年，我們重申先前1.35億美元至1.45億美元的營收預期，並將調整後EBITDA預期從-4200萬美元至-3800萬美元上調至-4000萬美元至-3600萬美元。我們仍將2023年第四季調整後EBITDA利潤率的目標設定為負兩位數。此外，我們將實現盈利的預期時間提前了一個季度。先前我們曾預計調整後EBITDA將在2024年底實現獲利，現在我們預計調整後EBITDA將在2024年第三季實現獲利。

Now I would like to turn the call over to the operator to open up the line for Q&A. Operator?

現在我想把電話轉給接線員，以便進行問答環節。接線生？

(Operator Instructions) Your first question comes from the line of Koji Ikeda with Bank of America.

（操作員說明）您的第一個問題來自池田浩二與美國銀行的合作。

Maybe a question for Kiwi or Mike. I wanted to really focus on this AI opportunity here. And thank you for all the color on where you guys might be going. But my question here is really about how to think about Cecilia and these other AI tools. These future AI tools that you're talking about is potential TAM expander. And the question is around AI tools. It should drive efficiency of the end market and it could lead to maybe thinking that it could result in less data that needs to be ingested onto the platform. So really just trying to understand the implications of how you think about these AI tools in the TAM.

或許可以問問 Kiwi 或 Mike。我想重點探討一下人工智慧帶來的機會。感謝你們分享了未來的發展方向。但我的問題其實是關於如何看待 Cecilia 和其他人工智慧工具。你們提到的這些未來人工智慧工具有可能拓展潛在市場規模 (TAM)。問題在於人工智慧工具本身。它應該能夠提高終端市場的效率，並可能減少平台需要接收的資料量。所以，我真正想了解的是，你們是如何看待這些人工智慧工具在潛在市場規模 (TAM) 中的作用的。

It's a great question. I think we're in early days of thinking about what the technology is going to do to different categories of spend. There are no doubt areas where the use of AI will be deflationary for example, in our core Review business. Indeed, the AI that we've baked into our products over the last 8 years has already been deflationary in the Review business, which is what allows us to deliver such material savings on a guaranteed basis to customers who switch from doing review using brute force armies of associates to AI-powered review using DISCO Review.

這是一個很好的問題。我認為我們目前還處於思考這項技術將如何影響不同類別支出的早期階段。毫無疑問，人工智慧的應用在某些領域會帶來通貨緊縮效應，例如我們核心的審核業務。事實上，過去八年來我們產品中整合的人工智慧技術已經在審核業務中實現了通貨緊縮，這使得我們能夠為那些從依靠大量員工進行繁瑣審核轉向使用DISCO Review人工智慧審核的客戶帶來顯著的成本節約。

If we go forward and think about the 4 categories of AI use that I talked about, one is question answering that Cecilia is an example of, another is document analysis. Third is drafting and document generation and the fourth is a natural language interface to the products I think you will see a mix of both taking existing categories of legal work and reducing spend on them by making them more efficient, but you will also see an expansion of the market for legal services because there will be an expansion of what is possible using these AI technologies.

如果我們繼續探討我之前提到的人工智慧的四大應用領域，一是問答（Cecilia 就是一個例子），二是文件分析，三是文件起草和生成，四是產品的自然語言介面。我認為你會看到，一方面，現有法律工作領域將得到改進，從而降低成本；另一方面，隨著人工智慧技術的拓展，法律服務市場也將隨之擴張。

To give you a very simple example, think about compliance, right? Every large bank, every large pharmaceutical company, every large defense contractor, any company that does a lot of M&A is faced with the issue of knowing whether or not there is problematic communication happening in e-mail, in documents, in Slack, in Bloomberg Chat, right?

舉個簡單的例子，想想合規性，對吧？每一家大型銀行、每一家大型製藥公司、每一家大型國防承包商，任何進行大量併購的公司都面臨著這樣一個問題：如何判斷電子郵件、文件、Slack、彭博聊天中是否存在有問題的溝通，對吧？

And the traditional approach of doing audits and having lawyers review samples of those documents can be replaced by an AI-powered system that is able not only to review those documents, but ideally provide actionable feedback at minimum to the legal department and the compliance department, but eventually possibly with the supervision of lawyers to the actual humans who are creating that communication in the first place. That's just one example of many that I could give. But I think we'll see with the rise of AI what we've seen with the rise of other technologies that it's deflationary in some areas, while at the same time, creating many more opportunities to use technology to create value for clients.

傳統的審計方式，即由律師審核文件樣本，可以被人工智慧系統所取代。該系統不僅能夠審核這些文件，理想情況下還能至少為法務部門和合規部門提供可操作的回饋，最終甚至可能在律師的監督下，為最初撰寫這些文件的人員提供回饋。這只是眾多例子中的一個。但我認為，隨著人工智慧的興起，我們將會看到，就像其他技術的興起一樣，它在某些領域會帶來通縮效應，但同時也創造更多利用科技為客戶創造價值的機會。

Got it. And just a follow-up here, if I may, for Michael. In your prepared remarks, you mentioned in early May, there was some adjustments in the operation that's leading to the higher adjusted EBITDA outlook. Could you dig into that a little bit more, please?

明白了。邁克爾，如果可以的話，我還想問一個後續問題。您在五月初的演講稿中提到，公司營運方面進行了一些調整，這導致了調整後 EBITDA 預期上調。能否再詳細解釋一下？

Well, so the higher adjusted EBITDA outlook and also pulling forward our adjusted EBITDA from exiting 2024 to Q3 of 2024 is related to -- but not exclusively related to. On May 9, we committed to a plan to reduce our workforce and representing about 47 employees and roughly 8% of our current global workforce. You'll see more details about this in the Q, but that's really related to pulling our guidance.

因此，上調調整後 EBITDA 預期，並將調整後 EBITDA 的確認時間從 2024 年底提前至 2024 年第三季度，與此相關，但並非完全相關。 5 月 9 日，我們承諾實施一項裁員計劃，涉及約 47 名員工，約占我們目前全球員工總數的 8%。您將在季度報告中看到更多相關細節，但這確實與我們事先確認業績指引有關。

Your next question comes from the line of Tyler Radke with Citigroup.

你的下一個問題來自花旗集團的泰勒拉德克。

This is [Kylie Tobin] on for Tyler Radke. And I just wanted to kind of piggyback on that last question. Can you -- with respect to those AI categories that you spoke about in the prepared remarks, which of these functionalities are you hearing customers specifically ask for the most? And as you guys look at your monetization strategy, which of these categories do you think provides the biggest opportunity for revenue uplift?

這裡是 Kylie Tobin 代 Tyler Radke 提問。我想就剛才的問題再補充一點。關於您在準備好的演講稿中提到的那些人工智慧類別，您能否具體說明一下，客戶最常要求哪些功能？另外，在你們的獲利策略中，您認為哪些類別最有可能帶來收入成長？

I think we're in an interesting moment where virtually every customer has been exposed to things like ChatGPT and many customers who have been exposed to things like Midjourney. And so you ask customers, where does your mind go to in terms of Generative AI, and you will get either My Category 1 or My Category 3. So you'll get either question answering or document generation. But I think there are opportunities to build really amazing products and consequently really amazing businesses in all 4 of the categories that I talked about. And I'll spend time, I think, focusing on the first one because that's where we have product experiences that we've put in front of customers and that, of course, is Cecelia.

我認為我們正處於一個有趣的時刻，幾乎所有客戶都接觸過 ChatGPT 之類的產品，許多客戶也接觸過 Midjourney 之類的產品。所以，如果你問客戶，他們對生成式人工智慧的印像是什麼，你會得到「我的類別 1」或「我的類別 3」的答案。也就是說，你會得到問答或文件產生。但我認為，在我提到的所有四個類別中，都有機會打造真正出色的產品，從而成就真正卓越的企業。我想我會花些時間重點談談第一個類別，因為我們已經向客戶展示了相關的產品體驗，而這個產品體驗當然就是 Cecelia。

If you think about what Cecilia lets customers do, it lets them ask questions in natural language and get back narrative answers in huge collections of private data. Our demonstration of this technology happens in the database populated with Enron's e-mails and documents, and you can ask questions like what was Enron's business in California or even ultimate questions like did Enron commit accounting fraud? You can follow up on those questions in natural language. And I think, importantly, for the legal use case, you can get back citations to documents in the database that support the narrative answer.

想想Cecilia能讓客戶做什麼，它允許客戶用自然語言提問，並在大量的私人數據中獲得敘事答案。我們示範這項技術時，使用的是包含安然公司電子郵件和文件的資料庫。您可以提出諸如「安然在加州的業務是什麼？」之類的問題，甚至可以提出更深入的問題，例如「安然是否犯有會計詐欺？」。您也可以用自然語言追問這些問題。而且我認為，對於法律應用場景而言，更重要的是，您可以獲得資料庫中支援敘述性答案的文件引用。

So that allows lawyers to get comfort that the answer is not a hallucination because they can go directly to the documents and those documents can then be the basis for further investigation in Cecelia or, again, importantly, in the rest of our platform. So one advantage of having not just the AI capabilities, but baking them into our platform is that after engaging in that Q&A with Cecilia, the lawyer can then learn from the documents that these other category of evidence is important, [full] adopt using search visualization collected from the client, make a production to the regulator of the court, the lawyer can seamlessly transition to deposition testimony and so on.

這樣一來，律師們就能確信答案並非臆想，因為他們可以直接查閱相關文件，而這些文件可以作為在Cecilia系統中或我們平台其他功能中進行進一步調查的基礎。因此，將人工智慧功能整合到我們的平台中，其優勢之一在於，律師在與Cecilia進行問答互動後，可以從文件中了解到其他類別的證據的重要性，並可充分利用從客戶處收集的搜尋可視化數據，向法院監管機構提交證據，律師還可以無縫接地進行庭前作證等等。

It's the power not just of a standalone AI feature, but incorporating those AI capabilities into a platform where lawyers are already doing their work. I think in the near term, Cecilia is the product that is closest to GA. We do think it will be material to customers, and we look forward to following it up with a fast sequence of releases across the other areas that I talked about.

它的優勢不僅在於獨立的AI功能，更在於將這些AI能力融入律師們日常工作的平台中。我認為在短期內，Cecilia是最接近正式發表的產品。我們相信它對客戶來說意義重大，我們也期待在接下來的幾個階段，也就是我之前提到的其他領域，推出一系列快速更新。

That's really helpful. And maybe one for Michael. The 2Q guidance, it assumes continued year-over-year revenue declines, but the full year implies a reacceleration in the business over the last few quarters. So how are you thinking about the back half of the year? And what kind of uplift are you anticipating from Cecelia or any of these new AI capabilities? Are they going to show up in FY '23 or will they potentially show up in FY '24?

這真的很有幫助。也許可以問問麥可。第二季業績指引假設營收年比持續下滑，但全年業績則顯示過去幾季業務有所回升。那麼，您對下半年的業績有何看法？您預計Cecelia或其他任何新的人工智慧功能會帶來怎樣的提升？它們會在2023財年推出，還是可能會在2024財年推出？

So Cecilia and these kind of newer technologies are not baked into our guidance for 2023. I do believe that technology is incredibly amazing. And I'll just mention that I'm old enough to be in the room that I recall, [pack in the dotcomer]. Many of the companies that were very successful that we all know of were start-ups. And one of the things that I think is most exciting about Cecilia is the fact that we have the data on our platform, and I think we and others in our position where we have the data, other companies and other industries can leverage this new technology and basically build on what we've been doing for years.

所以，Cecilia 和這類新興技術並沒有被納入我們 2023 年的發展規劃。但我確實認為科技非常了不起。順便提一下，我年紀夠大，還記得當年網路泡沫破滅的那段時期。我們熟知的許多成功企業最初都是新創公司。我認為 Cecilia 最令人興奮之處在於，我們在平台上擁有大量數據。我認為，像我們這樣擁有數據的公司，以及其他產業和企業，都可以利用這項新技術，在我們多年來累積的經驗基礎上繼續成長。

We've been doing AI for years. And Cecilia, I do believe, is going to be incredibly helpful for us and our clients. In terms of the guidance, for the back half of the year. It does obviously require reacceleration of our revenue. We are seeing in the pipeline and the funnel some positive signs, and there is our expectation and optimism that, that's going to turn into reacceleration of the revenue. And so we are optimistic on where we're going to end the year.

我們從事人工智慧領域多年。我相信，Cecilia 對我們和我們的客戶都將大有裨益。就下半年的業績展望而言，顯然需要我們重新加速營收成長。我們從潛在客戶和銷售管道中看到了一些積極的跡象，我們預期並樂觀地認為，這些跡象將轉化為營收的重新成長。因此，我們對年底的業績充滿信心。

Your next question comes from the line of Jackson Ader with SVB.

你的下一個問題來自傑克森·阿德與SVB的對話。

The first one is on Kiwi, your commentary on the large customers and larger matters, reducing their usage and their spend with you. Just curious -- are they reducing the usage and going to other providers that maybe have more of a legal service display with them or are there just not the level of activity or the level of matters that maybe existed a year or 2 ago?

第一個問題是關於Kiwi的，您提到大客戶和大型案件減少了使用和支出。我很好奇——他們是減少了使用量，轉而選擇其他提供更全面法律服務的供應商，還是僅僅因為業務量或案件數量不如一兩年前那麼多了？

No. I'm sure there may be a handful of exceptions. But what I'm talking about is not large customers defecting to the competition. It's just large customers reducing their usage on our platform. And to give you a sense of what I mean, there are things you can do on the margin of any case that would increase or decrease your usage of our platform. For example, if you are not closely managing the bottom line, you might just collect lots of data broadly from lots of witnesses, toss it into the platform. You get to the review when you get to the review.

不，我確信可能會有一些例外。但我說的不是大客戶流失到競爭對手那裡，而是大客戶減少了對我們平台的使用。為了讓您更理解我的意思，您可以做一些事情，這些事情可能會影響您對我們平台的使用量。例如，如果您並不特別在意最終結果，您可能只是從許多證人那裡廣泛收集大量數據，然後將它們上傳到平台。至於審查結果，您到時候再看吧。

Once you've conducted the review, you leave all the data up there as you go through the downstream parts of litigation like depositions, motions for summary judgment trial, you might even leave the data up during dependency of the appeal. It's not that you aren't doing anything with it, but you're coming in there occasionally over the course of sometimes a multiyear period as the litigation unfolds. So that was true for many of our large customers for many years, call it, for 10 years. What we've been seeing lately is as some of these legal departments come under pressure to provide cost savings as part of a broader company focus on profitability.

完成審查後，在訴訟的後續階段，例如取證、簡易判決動議審理，甚至在上訴階段，您都會保留所有資料。這並非意味著您完全不用這些數據，而是在訴訟進行過程中，有時會持續數年，偶爾需要使用它們。多年來，我們的許多大客戶都是這樣做的，比如說，長達十年。但最近我們發現，隨著公司整體獲利能力的提升，一些法務部門面臨降低成本的壓力。

For our large customers, DISCO'S line item can be material to them. And so it occurs to them to see what they can do to reduce the total cost of their DISCO usage. And so what they can do is trim various parts of that workflow that I talked about, instead of collecting data from every witness imaginable, maybe we can collect data from fewer witnesses instead of loading it immediately, maybe we can wait a month or 2 and load it closer to the discovery deadline. Instead of leaving it up there, maybe we can call it down so that we leave only the most relevant parts up there and archive the rest in a way where it can be restored, but isn't consuming active usage in DISCO.

對於我們的大客戶而言，DISCO 的費用對他們來說可能至關重要。因此，他們會考慮如何降低 DISCO 的總使用成本。他們可以精簡我之前提到的工作流程中的某些環節，例如，與其從所有可能的見證對象收集數據，不如減少收集的見證對象數量；與其立即加載數據，不如等待一兩個月，在接近取證截止日期時再加載；與其將數據一直保留在伺服器上，不如將其下架，只保留最相關的部分，並將其餘部分歸檔，以便佔用日後的資源

And maybe when the case is resolved rather than keep the data around because there is a prospect for follow-on related litigation, maybe we take it down, and we'll simply reload it later if we need to. Another piece that I talked about in my prepared remarks is, of course, the use of DISCO ECA, early case assessment, which is a set of capabilities that allows customers to load large volumes of data at a lower price point, explore and prioritize that data using tools like DISCO AI and search visualization and then promote data out of DISCO ECA into active review at a higher price point. We've seen an uptick in usage of DISCO ECA. Again, we think that's a phenomenon of customers, especially large customers being a little bit more careful about their volume of usage on our platform.

或許，當案件解決後，與其因為可能引發後續訴訟而保留數據，我們不如將其刪除，以後如有需要再重新上傳。我在準備的演講稿中也提到了DISCO ECA（早期案件評估）的使用。 DISCO ECA提供一系列功能，讓客戶以較低的價格加載大量數據，使用DISCO AI和搜尋視覺化等工具探索數據並確定優先級，然後以更高的價格將數據從DISCO ECA提升到正式審查階段。我們發現DISCO ECA的使用量增加。我們認為，這主要是因為客戶，尤其是大型客戶，更謹慎地控制他們在我們平台上的使用量。

Okay. All right. That was thorough. Michael, any kind of functional details that you can provide on the 47 people in the workforce reduction, whether it was sales or R&D or something like that?

好的。好的。這很全面。邁克爾，關於裁員的47人，你能提供一些具體的業務細節嗎？例如他們來自銷售、研發還是其他部門？

It was across all departments.

所有部門都出現了這種情況。

Your next question comes from the line of Derrick Wood with TD Cowen.

你的下一個問題來自德里克·伍德和TD·考恩的連線。

Great. Nice job on the rebound in new customers. I guess looking back to Q4, what do you think caused that blip? And give us a sense for what average deal sizes for new customers are trending versus maybe what it looked last year -- what it looked like last year? And then just, I guess, as you look at new customer generation looking forward, it sounds like you have good healthy pipelines being built. Do you see any kind of behavior where people are investigating ChatGPT or there are new start-ups that are talking about Generative AI capabilities and potential for a longer sales cycle given what's going on with Gen AI right now?

太棒了！新客戶數量回升動能良好。回顧第四季度，您認為是什麼原因導致了那次小幅下滑？能否介紹一下新客戶的平均交易規模趨勢，並與去年同期進行比較？展望未來，您認為您正在建立一個健康的客戶管道。您是否觀察到有人在研究 ChatGPT，或者有新的新創公司在探討生成式人工智慧 (Generative AI) 的功能，以及鑑於目前生成式人工智慧的發展趨勢，延長銷售週期的可能性？

Sure. So on customer count, we're delighted by the rebound to what has been sort of more historical rate of new customer additions in the quarter. I don't think there's really anything magical to it. If you break it apart into its components, what strengthened is churn, right? So deactivations improved quarter-over-quarter, contributing to the higher net add number, which is, of course, positive. Related to both of your questions, when we think about customers in their first year, some of these numbers can be a little bit complicated because until a customer gets to, say, 3 active matters or 6 active matters, it's possible for the customer to add a matter, hit 2 matters and then touch zero at some point before coming back.

當然。就客戶數量而言，我們很高興看到本季新增客戶數量回升至更接近歷史平均值。我認為這其中並沒有什麼特別神奇之處。如果將其分解開來，流失率的下降才是關鍵，對吧？因此，客戶停用率環比有所改善，從而促成​​了淨新增客戶數量的成長，這當然是件好事。關於您提出的兩個問題，當我們考慮客戶第一年的情況時，一些數據可能會比較複雜，因為在客戶達到一定數量的活躍事項之前，例如3個或6個，客戶有可能添加一個事項，達到2個，然後再次降至零，之後又重新開始。

So you have some of that bouncing around in the numbers. It also complicates thinking about what is the average deal size because when you sign up the customer, they simply sign a contract that has unit pricing. And then the size of that deal will be determined by how much data is involved in the first case or a couple of cases that they put on the DISCO platform. But zooming out a little bit kind of in the aggregate, think about customers starting out in the middle 5 figures kind of spend and then growing over time to low 6 figures, middle and upper 6 figures and then our biggest customers winding up in the low 7 figures.

所以，這些數字裡會有一些波動。這也使得平均交易規模的計算變得複雜，因為客戶在簽約時只需簽署一份包含單價的合約。而最終的交易規模則取決於他們在DISCO平台上提交的第一個或幾個案例的資料量。但如果從整體來看，客戶最初的消費額在五位數左右，然後隨著時間的推移增長到六位數出頭，再到六位數中段和高段，而我們最大的客戶最終消費額會達到七位數出頭。

And again, that journey historically took sort of 3 to 5 years in more recent times, sometimes that journey has been accelerated for logos, especially when there is a repeat buyer at the logo, meaning someone who went through a DISCO adoption journey at a prior company switches companies and then adopt DISCO at the new company. I don't think we've seen any material trend in terms of the size of first year spend.

而且，從歷史上看，這個過程通常需要3到5年。近年來，對於一些品牌來說，這個過程有時會加快，尤其是當品牌有回頭客時，也就是說，之前在一家公司經歷過DISCO系統採用流程的客戶跳槽到另一家公司後，又在新公司繼續採用DISCO系統。我認為，就第一年的支出規模而言，我們還沒有看到任何實質的趨勢。

Okay. Just the follow up on the second question around how often is ChatGPT coming into conversations. I think you mentioned that every customer is aware of it now, every company is. I mean, is that creating some part of the sales cycle on a conversation? And are there any different vendors showing up in deal cycles?

好的。關於第二個問題，也就是 ChatGPT 在對話中出現的頻率，我想再補充一點。您提到現在每個客戶、每家公司都知道它的存在。我的意思是，它是否在對話中構成銷售流程的一部分？在交易週期中，是否會有不同的供應商出現？

So far, it's been an accelerator. We were, of course, delighted to be able to announce Cecelia our question answering service for Ediscovery at Legal Week a couple of months ago in March. We had tremendous reaction at the show. We had a big event at the end of one of the days where we had hundreds of people come in and did a live unveiling. Of course, lots of one-on-one meetings following up with customers then and lots of interest from customers in the months that have followed so tremendous demand there.

到目前為止，它確實起到了加速作用。當然，我們非常高興能在幾個月前的三月的法律週上發布我們的電子取證問答服務Cecelia。展會上反應熱烈。其中一天結束時，我們舉辦了一場大型活動，數百人前來現場觀看。當然，之後我們也與許多客戶進行了後續的一對一會談，並在接下來的幾個月裡收到了大量客戶的關注，可見市場需求非常旺盛。

I think it's been a door opener, both for customers, new and existing, who are interested in Cecelia, but also it's just an opportunity to go talk about the rest of the DISCO product portfolio, so a form of lead generation. We haven't seen, I think, much change in Ediscovery in the competitive set based on GPT so far. And we've seen even less impact in our other product areas. But if I were a predicting person, I would say that we will see competition both from incumbents and presumably from start-ups in both Ediscovery and review likely first, and we'll see other AI applications in other areas. In terms of facing that competition, I think DISCO is in a very good position because, of course, we've been doing AI for 8 years now.

我認為這為新舊客戶打開了一扇門，讓他們有機會了解Cecelia，同時也為我們提供了一個機會，可以進一步介紹DISCO的其他產品組合，從而有效地擴展客戶。就目前而言，基於GPT的電子取證競爭格局尚未發生太大變化。而對我們其他產品領域的影響則較小。但如果讓我預測未來，我會說，電子取證和審查領域將會率先面臨現有企業和新創公司的競爭，而其他領域也將出現人工智慧應用。面對這些競爭，我認為DISCO處於非常有利的地位，因為我們在人工智慧領域已經深耕八年了。

And as Michael pointed out, we have the advantage of having access not only to a large collection of private data that is very hard for other people to obtain, but also lawyers are doing their work in our platform. And so we get lots of what we used to call training data. It's observations of how layers interact with label and annotate the data in our platform, all of which can be used to train these models. I think another thing that you're seeing with some of these AI releases is how they're compounding.

正如邁克爾指出的那樣，我們的優勢在於不僅能夠訪問大量其他人很難獲得的私有數據，而且律師們也在我們的平台上開展工作。因此，我們獲得了大量我們過去稱之為訓練資料的資訊。這些數據包括觀察各層如何互動、標記和標註我們平台上的數據，所有這些都可以用來訓練這些模型。我認為，在這些人工智慧產品發布過程中，我們也可以看到它們是如何相互疊加的。

So the company that releases the first great chatbot gets not only market share, but they also get all the signal that comes from people interacting what questions are they asking? What follow-ups do they ask? Which questions do they like or dislike? And so we think that's another benefit of being early to market with products like Cecilia, it's not only grabbing share early on, but also grabbing an unfair share of the training data that can be used to power subsequent AI releases.

因此，率先推出優秀聊天機器人的公司不僅能獲得市場份額，還能獲得用戶互動產生的所有訊號：他們會問什麼問題？會追問什麼？他們喜歡或不喜歡哪些問題？我們認為，像 Cecilia 這樣的產品能夠搶佔市場先機，這本身就是一大優勢。它不僅能搶佔早期市場份額，還能獲得大量訓練數據，這些數據可用於支援後續人工智慧產品的發布。

Great. Well, excited to see the new product come to market later this year. Thanks.

太好了。非常期待這款新產品今年稍後上市。謝謝。

Your next question comes from the line of Arvind Ramnani with Piper Sandler.

你的下一個問題來自 Arvind Ramnani 和 Piper Sandler 的那段對話。

I wanted to kind of just go back to kind of the Review business, right? Like, I mean, if you have any update on that? I know there was fairly volatile. Has that kind of stabilized? And what's your kind of outlook on the Review business?

我想回歸評論業務，對吧？我是說，你這方面有最新進展嗎？我知道之前市場波動比較大。現在穩定下來了嗎？你對評論業務的未來前景有什麼看法？

I do have a cautious update to share about the Review business. So first, finally, it is up quarter-over-quarter, which is good, although modestly. Second, this is a quarter where -- so not only are all the big ticket reviews gone, but the tail of all those big ticket reviews is completely out of the number. And then additionally, we talked about improvement in the pipeline metrics that we look at.

關於評論業務，我有一些謹慎的最新消息要分享。首先，終於實現了環比成長，雖然增幅不大，但總體來說是個好消息。其次，本季不僅所有高價評論都已取消，而且這些高價評論的尾部也完全不計入統計數據。此外，我們也討論了我們關注的通路指標有所改善。

So how many opportunities are there, how many wins are there, what are the win rates and so on. And I can tell you, given where we are today, we are seeing finally an uptick in the Review business. I want to [care] that optimism with caution. We will report on Q2 results when we report on Q2 results. But I think 2 good facts are it seems like we're back in the state of growth in that business and all of the big ticket items and their tails are fully out.

那麼，機會有多少？成功案例有多少？勝率是多少？等等。我可以告訴大家，就我們目前的狀況而言，評論業務終於出現了回暖跡象。但我對這種樂觀情緒持謹慎態度。我們會在公佈第二季業績時再做報告。不過，我認為有兩個值得關注的事實：一是該業務似乎已經重回成長軌道；二是所有大額項目及其後續階段都已全面展開。

Terrific. And then you certainly provided a lot of color in your prepared remarks and the past Q&A on AI. But I didn't want to kind of think about how should we think about it, right? Like I mean there's one part -- one way of looking at this and saying, like, hey, you have Cecilia and to be an AI product-driven company from the beginning, and this is not like a fact that you're chasing, right?

太棒了。您在事先準備好的發言稿和先前關於人工智慧的問答環節中確實提供了許多精彩的見解。但我不想去思考我們該如何看待這個問題，對吧？我的意思是，看待這個問題的一種方式是，比如說，你們有Cecilia，從一開始就是一家以人工智慧產品為驅動的公司，這並不是你們刻意追求的目標，對吧？

So it feels like you're in a unique position to take advantage of this broader developments in AI technology. But on the other hand, what you also mentioned as part of Q&A is like there's going to be these incumbents, either like startups or other companies kind of -- that have been in the business kind of really sharpening the product with AI. Of course, you'll have -- you did clarify you have a large data set and a unique data set that helps you train things a lot better. But is this still like a mix thing? We have to see like -- is it mixed -- AI is mixed for you all or do you really look at it as an opportunity or is there a potential for being a threat as well?

所以感覺你們在利用人工智慧技術的廣泛發展方面處於獨特的優勢地位。但另一方面，您在問答環節也提到，一些老牌企業，例如新創公司或其他一些公司，一直在利用人工智慧技術不斷改進產品。當然，你們也擁有──你們也提到你們擁有龐大的資料集和獨特的資料集，這有助於你們更好地訓練模型。但這仍然是一種混合局面嗎？我們需要看看——對你們來說，人工智慧究竟是混合的，還是混合的？你們真的把它看作是個機遇，還是也存在著潛在的威脅？

I think AI is clearly an opportunity. And I can tell you, it's not just the product things, but it's -- I'm not sure if I put this in my prepared remarks. But for me, I have not been as excited about any fundamental technology advance since the mid-90s web, which is -- I was in school at the time, that's what convinced me to study computer science. And we talked a bit about how we're going to bake some of these AI technologies into the products. I can also share with you that throughout DISCO internally, we have launched an aggressive campaign to take every part of our business and use AI.

我認為人工智慧顯然是一個機會。我可以告訴你們，這不僅關乎產品，還有──我不確定我是否在準備的發言稿中提到過這一點。但對我來說，自從90年代中期出現網路以來，我還沒有對任何一項基礎技術進步感到如此興奮。當時我還在上學，正是網路讓我下定決心學習電腦科學。我們之前也討論過如何將一些人工智慧技術融入產品中。我也可以告訴你們，在DISCO內部，我們已經發起了一項積極的行動，將人工智慧應用於我們業務的各個環節。

Everything from drafting, communications and internal documents to -- we've trained internal chatbots across the entire knowledge base and ticket responses so that we're able to give people first drafts of what they should be saying to customers and answer internal questions about how our products and systems work. And I think that's just the beginning. So I think AI is a huge opportunity. I do think you're right, though, that part of AI being such a huge opportunity is that there are large categories of legal work, including some of the work that we either accelerate or do that are going to be largely automated as AI is better productized and becomes even more effective.

從起草文件、溝通和內部文檔，到——我們已經訓練了涵蓋整個知識庫和工單回复的內部聊天機器人，這樣我們就能為員工提供與客戶溝通的初稿，並解答內部關於我們產品和系統如何運作的問題。我認為這只是個開始。所以我認為人工智慧蘊藏著巨大的機會。不過，你說的也很有道理，人工智慧之所以蘊藏如此巨大的機遇，部分原因在於，許多法律工作領域，包括我們正在加速推進或正在進行的一些工作，隨著人工智慧產品化程度的提高和效能的增強，將會實現高度自動化。

To give you an example, I don't believe that in 10 years, there will be a lot of lawyers reading documents. That is going to be obsoleted using AI. I would, of course, believe that DISCO is in the best position to do that obsoleting. And so the way I might phrase this, I think we're a little bit like Netflix, when Netflix made the shift from DVDs in the mail to streaming. We are well-positioned to go take advantage of this opportunity. But if we do not take advantage of this opportunity, then I think our prospects wouldn't be great.

舉個例子，我不認為十年後還會有很多律師在閱讀文件。人工智慧會取代這項工作。當然，我認為DISCO最有能力完成這項變革。所以，我這麼說吧，我覺得我們有點像Netflix，就像Netflix當年從DVD郵寄轉型到串流播放一樣。我們完全有能力抓住這個機會。但如果我們不抓住這個機會，我認為我們的前景並不樂觀。

Your next question comes from the line of Brent Thill with Jefferies.

你的下一個問題來自 Jefferies 的 Brent Thill。

This is Luv Sodha on for Brent Thill. Maybe first for Kiwi, I know in your prepared remarks, you talked a little bit about the fact that revenue might not grow in line with the pipeline improvement that you've seen. I guess going forward, is this more of a near-term trend or should we expect this to be more of a longer-term trend? And if it isn't a near-term trend and what gives you confidence that it will rebound more in the longer term?

這裡是 Luv Sodha，替 Brent Thill 發言。首先，我想問 Kiwi，您在事先準備好的演講稿中提到，收入成長可能不會與您看到的項目儲備改善同步。我想問的是，這更像是短期趨勢，還是應該被視為長期趨勢？如果不是短期趨勢，您又是如何確信它在長期內會反彈的呢？

So the only thing I can do is to report on what we're seeing in the market as we see it. And what I tried to convey in my prepared remarks is what we are hearing from our largest customers, and it's -- there's nothing fancy about it. It's -- we are at companies where suddenly profitability and EBITDA matters a lot more. And we've received just like every other department, a mandate from finance saying we would like you to cut budgets by X percent. And when DISCO is a small line item, sometimes that mandate misses us, right?

所以我唯一能做的就是如實報告我們目前在市場上的觀察。我在事先準備好的發言稿中試圖傳達的是我們從最大客戶那裡聽到的回饋，其實沒什麼特別的。我們身處的這些公司，突然間獲利能力和 EBITDA 變得至關重要。和其他部門一樣，我們也收到了財務部門的指令，要求我們削減 X% 的預算。而當 DISCO 只是很小一部分預算時，我們有時就會忽略這項指令，對吧？

But when DISCO is a big line item as it is at our largest clients, then the client calls us and says, hey, we've received this mandate from finance. How can we help -- how can you help, rather? And our first response to them is always that the way to reduce legal spend is to buy more DISCO. And sometimes that works, it really does. In other times, we get further along in that conversation, and we start either negotiating total spend or thinking about ways that usage can be reduced in that discussion with our large clients. We try to ensure there is something for both sides.

但當 DISCO 費用像我們最大的客戶一樣成為一項大額支出時，客戶就會打電話給我們說：「財務部門給我們下達了這項指令，我們能幫上什麼忙？」——或者更確切地說，你們能幫上什麼忙？我們通常的第一個反應是，降低法律支出的方法就是增加 DISCO 的使用量。有時候這招確實管用。但有時候，隨著討論的深入，我們會開始與大客戶協商總支出，或探討如何降低使用量。我們努力確保雙方都能從中受益。

And maybe we help the client achieve its profitability or cost-cutting goals in the near term in exchange for longer-term commitments or an agreement to roll out a (inaudible) swathe of DISCO products. But those kinds of conversations from our largest clients, were simply not conversations that we were having during the 10-plus years in our history of economic boom times, and they do present a headwind to growth. In terms of whether they will continue, I don't have anything special to add other than I have faith in the U.S. economy and the world economy.

或許我們會幫助客戶在短期內實現獲利或成本削減目標，以此換取更長期的承諾，或達成協議，推廣一系列（聽不清楚）DISCO產品。但與我們最大的客戶進行的這類對話，在過去十多年的經濟繁榮時期，我們從未遇到過，而且它們確實對成長構成了阻力。至於這種情況是否會持續下去，我沒有特別要補充的，只是我對美國經濟和世界經濟充滿信心。

And at some point, we won't be in such grimly perceived economic times. And when that happens, I don't expect that there will be any fundamental change in the trends that show increased exposure to legal issues for companies operating around the world and increased volumes of data because of the rise of not only e-mail and documents, but also application data, things like chat streams, application data like JIRA, Slack, Salesforce, video, audio and so on. So I guess what I'm saying is I'm a believer in those same long-term trends as opposed to some questions we seem to be getting from customers that are driven by a short-term desire to improve profitability.

總有一天，我們不會再面臨如此嚴峻的經濟狀況。即便如此，我也不認為企業面臨的法律風險增加以及資料量激增（不僅包括電子郵件和文檔，還包括聊天記錄、JIRA、Slack、Salesforce 等應用資料、視訊、音訊等等）等趨勢會發生根本性改變。所以，我想說的是，我更看好這些長期趨勢，而不是像某些客戶一樣，只是出於短期獲利需求而提出問題。

Got it. Perfect. And one quick follow-up for Michael, if I may. Just a quick question on -- given the beat this quarter and you're not flowing through the beat into the full year guide. I guess -- and the fact that comps keep getting easier sequentially. Could you just give us some color into what's being baked into the guidance for the full year?

明白了，太好了。如果可以的話，我還有一個問題想問麥可。鑑於本季業績超預期，但您並沒有將此預期延續到全年業績指引中。我猜——而且同比數據逐週下降——您能否詳細說明一下全年業績指引中包含了哪些因素？

Yes. So we -- in terms of the back half of the year, it obviously does require reacceleration of the revenue. We do look at, and as Kiwi mentioned, at our metrics every week, every day and our pipeline review metrics, our funnel metrics. And we are seeing positive signs. We have seen some positive signs that we believe will lead to the revenue growth and the reacceleration of the revenue that we need to hit our numbers for the full year.

是的。所以，就下半年而言，顯然需要重新加速營收成長。正如Kiwi所提到的，我們每週、每天都會關注各項指標，包括銷售管道指標和轉換漏斗指標。我們看到了一些積極的跡象。我們相信，這些正面跡象將有助於實現營收成長，並重新加速營收成長，從而達成全年目標。

Got it. But just to be clear, are you baking in more conservatism now than you did, say, 3 months ago?

明白了。但為了確認一下，你現在是否比三個月前更保守了？

No.

不。

Your next question comes from the line of Scott Berg with Needham.

你的下一個問題來自 Scott Berg 與 Needham 的對話。

This is Rob Morelli on for Scott. Understanding it might be too early to have a concrete answer here. But last quarter, you mentioned a push towards corporate versus selling it to law firms. Do you have any insight into that push and how it's fared so far (inaudible)?

這裡是羅伯·莫雷利，替斯科特回答。我知道現在給出確切答案可能還為時過早。但上個季度，您提到要將業務重心轉向企業客戶，而不是賣給律師事務所。您對此舉措有何見解，以及目前的進展如何？ （聽不清楚）

Yes, that's terrific. That's a shining point, I think, for DISCO. We shared some numbers in our annual earnings calls of the last earnings call about total number of corporate direct, logos. And I believe we also shared the growth of that on a year-on-year basis, which exceeded overall logo growth substantially. So this has been a strategy of ours over the long term. And it is, of course, since the beginning, the dollars that have flown into DISCO as revenue have always come from the budgets of corporate legal departments.

是的，太棒了。我認為這是DISCO的一大亮點。我們在上次財報電話會議上分享了一些關於企業直接客戶總數的數據。我相信我們也分享了該數字的同比增長情況，其增速遠超整體客戶數量的增長。因此，這始終是我們長期奉行的策略。當然，自公司成立以來，流入DISCO的收入一直都來自企業法務部門的預算。

But earlier in our journey as a company, we were much more dependent on law firms as a channel to go reach the [open-end] decision-maker in the legal department of large enterprises. And as the company has grown and especially following our IPO, I think we've been viewed by many of those customers as a more credible partner to work with directly and perhaps a partner that they have more confidence will be around for the long haul. In any event, whatever the reason we've seen substantial success in being able to forge those relationships directly. We have a constituted enterprise sales group that is maniacally focused on growing and forming new corporate direct relationships. We think it's very much part of the future.

但在我們公司發展的早期，我們更依賴律師事務所作為管道，才能接觸到大型企業法務部門的決策者。隨著公司的發展，尤其是在上市之後，我認為許多客戶已經將我們視為更值得信賴的合作夥伴，他們也更有信心與我們長期合作。無論如何，無論出於何種原因，我們都已在直接建立這些關係方面取得了顯著的成功。我們擁有一支專業的企業銷售團隊，專注於拓展和建立新的企業直接合作關係。我們認為這才是未來的發展方向。

Got it. And then understanding Cecilia is not looped into the updated guidance. As you look across your product portfolio, what sales or use of modules are you anticipating will drive another revenue growth acceleration that's implied in your updated guidance?

明白了。但是，Cecilia 的情況並沒有納入更新後的績效指引中。縱觀您的產品組合，您預計哪些模組的銷售或使用情況將推動更新後的業績指引中提到的另一輪營收成長加速？

Well, our 2 -- so the -- our original product, DISCO Ediscovery, still represents the vast, vast majority of revenue. And so growth in that product line is the lion's share of overall growth. I think we are having growing optimism about review contributing to growth, although that, again, you have to let all of the comps roll off. And this quarter, Q1 is the first quarter where all of the big ticket review at its tail has fully dropped out. So I do think review can be a contributor. And then the newer products, we think can be material percentage contributors of growth but because of the scale of those businesses, they're not likely to be big movers of overall top line growth. So Ediscovery, review and a dash of everything else.

嗯，我們的第二款產品——也就是我們最初的產品DISCO電子取證——仍然佔據了絕大部分的收入。因此，該產品線的成長是整體成長的主要動力。我認為我們對審查業務的成長貢獻越來越樂觀，儘管如此，我們仍然需要剔除所有同店銷售額的影響。而且，本季（第一季）是所有尾單的大額審查業務完全退出後的第一季。所以我認為審查業務可以成為成長的貢獻者。至於新產品，我們認為它們也能貢獻相當可觀的成長份額，但由於這些業務的規模，它們不太可能成為整體營收成長的主要推動力。所以，電子取證、審查業務以及其他一些業務共同構成了我們整體的成長。

Your next question comes from the line of Parker Lane with Stifel.

你的下一個問題來自帕克巷與史蒂費爾的合作系列。

This is Matthew Kikkert on for Parker. I'd love to touch on the new office you opened in India. What kind of incremental impact do you think that offering 24/7 customer support will be able to give you? Secondly, how quickly do you believe you'll see a return on this investment? And do you foresee any other international growth this year.

這裡是帕克公司的馬修·基克特。我想談談您在印度開設的新辦事處。您認為提供全天候客戶支援能為公司帶來哪些方面的提升？其次，您預計這項投資多久才能回本？您預計今年還會有其他國際業務成長嗎？

We're very excited about the launch of the Delhi office. And we are treating it -- many companies in our space and of course, in many spaces, have treated their India presences as low-cost expansion -- the term we tried to avoid is offshoring. We don't think about the Delhi office that way. We think there is the same caliber of talent available in Delhi as there is in London and New York. And over time, we intend to build the Delhi office into a full-fledged office across essentially every department of our business. And we've made great strides towards accomplishing that already and intend to continue growing that office over the course of the year.

我們對德里辦事處的成立感到非常興奮。我們對待德里辦事處的態度截然不同——許多同行業的公司，當然也包括其他許多行業的公司，都將在印度的業務拓展視為低成本擴張——我們盡量避免使用「離岸外包」這個詞。我們並不這樣看待德里辦事處。我們認為德里擁有與倫敦和紐約同等級的人才。隨著時間的推移，我們計劃將德里辦事處發展成為一個涵蓋我們業務幾乎所有部門的成熟辦事處。目前，我們已朝著這個目標取得了長足的進步，並計劃在今年繼續發展壯大德里辦事處。

I think one of our learnings from the experience during COVID hasn't been that we really like remote work, but it has been that we can have these multiple centers of excellence in a kind of a global nexus of talent, Austin, New York, London and then Delhi. So that's how we think about it. We think that globalization gets us access to better talent and obviously also has the effect of improving our profitability profile. We have nothing else to announce on the globalization front right now, but I think we are open to the idea of announcing a similar expansion in the future. Some of that, of course, will depend on the overall rate of head count expansion in the business. In terms of 24/7 sport, we've always provided 24/7 support, but probably Delhi makes the lives of the Americans a little bit better since they don't have to stay up at night.

我認為我們從新冠疫情期間的經驗中學到的一點並非我們真的喜歡遠距辦公，而是我們可以建立多個卓越中心，形成一個全球人才樞紐，例如奧斯汀、紐約、倫敦和德里。這就是我們的思考方式。我們認為全球化讓我們能夠接觸到更優秀的人才，並且顯然也有助於提升我們的獲利能力。目前我們在全球化方面沒有其他消息可以公佈，但我認為我們未來會考慮類似的擴張計劃。當然，這在一定程度上取決於公司整體的員工成長速度。就全天候體育賽事服務而言，我們一直提供全天候支持，但德里中心的設立或許能讓美國員工的生活更輕鬆一些，因為他們不必熬夜了。

Your next question comes from the line of David Hynes with Canaccord Genuity.

你的下一個問題來自 Canaccord Genuity 的 David Hynes。

So Kiwi, look, in your answer to both Jackson's and Luv's questions, you talked about how and why a core Ediscovery customer may be looking to trim their spend. But I think when you put a comment out there without any quantification it's going to make investors nervous, right? So is there any way you can put some bands around the magnitude of trimming that you're seeing? I mean is it 5%? Is it 10%? Is it more or less? Like anything to help us think about how that impacts the numbers would be super helpful.

Kiwi，你看，你在回答Jackson和Luv的問題時都提到了核心電子取證客戶可能正在考慮削減開支的原因和方式。但我覺得，如果你只是發表一些沒有量化數據的評論，會讓投資人感到不安，對吧？所以，你能否大致估算一下你觀察到的削減幅度？比如，是5%？ 10%？還是更多或更少？任何有助於我們思考這會對財務數據產生何種影響的資訊都將非常有幫助。

We'll take that as good feedback. We don't have anything for you now. I can give you not in the aggregate, but I'm involved in most of these conversations, so I can sort of anecdotally share with you. Sometimes it can be a little bit -- you have to disambiguate just the ordinary ups and downs of usage and then the sort of keep usage -- keep the legal matters fixed, but reduce the amount of usage of the DISCO platform on those legal matters. Again, don't hold us to these exact numbers. But sort of color wise, I would put it in the kind of 10%, 15%-ish range for a big customer, but I want to wrap that in all the disclaimers that can be given. That's purely anecdotal.

我們會把這當作很好的回饋。目前我們沒有什麼具體數據可以提供給您。雖然我無法提供整體數據，但我參與了大部分相關討論，所以可以分享一些個人經驗。有時情況可能有點複雜—您需要區分正常的使用量波動，以及為了解決法律問題而減少DISCO平台使用量的情況。再次強調，請不要把這些數字當作確切的參考。但大致來說，對於大客戶而言，使用率大概在10%到15%左右，但我必須在此聲明所有可能的免責聲明。這純粹是個人經驗。

Yes. Understood. Okay. And then Michael, a follow-up for you. So the guidance implies that OpEx goes down in Q2. I presume based on the timing of the restructuring you talked about today, OpEx probably goes down a little bit more in Q3. From there, like what kind of OpEx growth are you embedding in the assumption that you can get to breakeven by Q3 '24? Like, does OpEx just kind of stay flat at that Q3 level or how are you thinking about it?

是的，明白了。好的。邁克爾，我還有一個後續問題。根據業績指引，營運支出預計在第二季下降。我根據你今天提到的重組時間推測，營運支出在第三季可能會進一步下降。那麼，假設你能在2024年第三季達到損益平衡，你對營運支出的成長預期是怎麼樣的呢？營運支出會維持在第三季的水準不變，還是你另有打算？

I mean so there's -- I mean, we've talked about this before, there's 4 levers to us getting to path to profitability. I will tell you that we're very committed to the path to profitability. And in my prepared remarks, we talked about where we expect EBITDA will be in Q4 of this year. We also provided color on the full year adjusted EBITDA numbers that we're targeting and also hitting adjusted EBITDA positive in Q3. There's 4 ways to get there. One is revenue growth, 2 is COGS and optimization and how we kind of work with AWS and the other vendors that hit COGS and also obviously, the below-the-line AWS cost.

我的意思是，我們之前也討論過，實現獲利有四個關鍵因素。我可以明確地告訴大家，我們致力於獲利。在我準備的發言稿中，我們談到了今年第四季 EBITDA 的預期目標。我們也詳細說明了全年調整後 EBITDA 的目標，以及第三季調整後 EBITDA 轉正的目標。實現盈利有四個途徑。一是營收成長；二是降低銷售成本 (COGS) 並進行最佳化，包括我們如何與 AWS 和其他影響銷售成本的供應商合作；當然，還有 AWS 的離線成本。

There's non-staff costs, just like we have some customers that Kiwi talked about who are approaching us to manage their total spend on the platform. I can tell you that we are doing the same thing with all of our vendors and partners to optimize and manage our overall non-staff, a lot of its software spend. And then obviously, our overall globalization effort and the New Delhi office is obviously part of our plan. So there's a big mix in how we're going to get there, but I will tell you we're very committed to getting there.

除了人員成本之外，還有一些其他成本，就像Kiwi提到的那些客戶一樣，他們正在聯繫我們，希望我們幫助他們管理在平台上的總支出。我可以告訴大家，我們也在與所有供應商和合作夥伴合作，優化和管理我們的整體非人員成本，其中大部分是軟體支出。當然，我們的全球化策略以及新德里辦事處也是我們計劃的一部分。所以，實現目標的方式多種多樣，但我可以肯定地說，我們致力於達成目標。

And your final question comes from the line of Mark Schappel with Loop Capital Markets.

最後一個問題來自 Loop Capital Markets 的 Mark Schappel。

Just going back to Cecilia. I was wondering if -- you talked a lot about Cecilia from a product perspective. Just wondering if you could talk about it from a go-to-market approach or maybe just talk a little bit about some of the activities that you have planned or you're currently undergoing with respect to your sales force to bring them up to speed on the new product and capabilities.

回到Cecilia的話題。我想問一下——您之前從產品角度談了很多關於Cecilia的內容。我想請您從市場推廣的角度談談，或者簡單介紹一下您為幫助銷售團隊快速了解新產品及其功能而製定或正在進行的一些活動。

Sure. Let me talk about it at a couple of different levels. So first, company-wide at DISCO, I kicked this off with an all hands some time ago, announcing not only our intent to bake these modern AI technologies into our products, but as I talked a bit about our intent to bake these modern AI technologies into the way we work in every part of our company. So there is an ongoing campaign of education so that you can pick an SDR at DISCO or someone working in business intelligence or someone working in engineering and ask them what's going on with large language models and what's an embedding and so they can tell you that and they can tell you how those technologies can be applied to make there and their colleagues work more efficient.

當然。讓我從幾個不同層面來談談。首先，在DISCO公司內部，我之前召開了一次全員大會，正式啟動了這項工作。在大會上，我不僅宣布了我們將把這些現代人工智慧技術融入我們的產品中，而且正如我之前提到的，我們還計劃將這些現代人工智慧技術融入到公司各個部門的工作方式中。因此，我們正在進行一項持續的培訓活動，您可以隨便找一位DISCO的銷售開發代表（SDR）、商業智慧部門的員工或工程部門的員工，問問他們大型語言模型是什麼，什麼是嵌入，他們就能為您解答，並告訴您這些技術如何應用到他們和同事的工作中，從而提高工作效率。

So at the biggest picture level, that's going on. In terms of the rollout we synchronized the announcement with Legal Week this year, a couple of months ago. At Legal Week, we did a series of invitation based suite meetings as well as a big party where we did public demo as well as hourly demonstrations at the booth. Since then, we've done a series of meetings with prospects and select new customers where we give them the opportunity to learn about the technology and to play with the product directly, right?

所以從整體來看，情況就是這樣。在推廣方面，我們今年將發表會與法律週（Legal Week）同步進行，就在幾個月前。在法律週期間，我們舉辦了一系列邀請制的商務會議，以及一場大型派對，並在派對上進行了公開演示，同時還在展位上提供了每小時一次的演示。此後，我們與潛在客戶和部分新客戶進行了一系列會議，讓他們有機會了解這項技術並直接體驗產品，對吧？

So to ask questions of Cecelia directly, gauge the answers, form the review about how applicable it would be to their practice. We've also done webinar-style presentations to broader audiences. And next week, we'll be at CLOC Legal Operations Conference, one of the larger conferences in the space. So there's that kind of awareness building and frankly, demand-building going on in the field. We have also enabled the sales team and the solutions architect team to be able to have these conversations with clients, and we've enabled a number of our partners to amplify the messaging around Cecilia.

因此，我們直接向Cecelia提問，了解她的回答，並據此評估她的方案對他們業務的適用性。我們也為更廣大的受眾舉辦了網路研討會形式的示範。下週，我們將參加CLOC法律營運大會，這是業界規模最大的會議之一。所以，我們正在努力提高公眾意識，坦白說，也積極創造市場需求。我們也讓銷售團隊和解決方案架構師團隊能夠與客戶進行這些對話，並讓一些合作夥伴擴大Cecilia的影響力。

As we get closer -- well, there's going to be a phased rollout where we're going to expose our largest existing clients to Cecilia, allowing them to use it on their own matters and collecting important early feedback, and then that will be followed by general availability. We are still working on the commercial terms that will surround general availability, and we'll have more to announce as we get closer to that time.

隨著發布日期的臨近，我們將採取分階段推廣的方式。首先，我們會向現有的最大客戶開放 Cecilia 的使用權限，讓他們在處理自身事務時進行測試，並收集重要的早期回饋。之後，我們將正式向所有客戶推出 Cecilia。目前，我們仍在討論正式發布相關的商業條款，並將在接近發布日期時公佈更多資訊。

And there are no further questions at this time. I will turn the call to Kiwi Camara, Co-Founder and CEO, for closing remarks.

目前沒有其他問題了。接下來我將把發言權交給共同創辦人兼執行長基維·卡馬拉，請他作總結發言。

Thank you for joining us today, and thank you for your interest in DISCO.

感謝您今天蒞臨現場，也感謝您對 DISCO 的關注。

This concludes today's conference call. You may now disconnect your lines

今天的電話會議到此結束。您可以斷開線路了。