使用警語:中文譯文來源為 AI 翻譯,僅供參考,實際內容請以英文原文為主
Operator
Operator
Greetings, and welcome to the BigBear.ai Holdings, Inc. second quarter 2026 earnings call and webcast. (Operator Instructions) As a reminder, this conference is being recorded.
各位好,歡迎收聽 BigBear.ai Holdings, Inc. 2026 年第二季財報電話會議與網路直播。(接線員指示) 提醒您,本次會議將被錄音。
It is now my pleasure to introduce your host, Sean Ricker, Chief Financial Officer. Thank you. You may begin.
現在很榮幸為各位介紹本次會議主持人:財務長 Sean Ricker。謝謝。您可以開始了。
Sean Ricker - Acting Chief Financial Officer
Sean Ricker - Acting Chief Financial Officer
Good afternoon, and thank you all for joining us today for our second quarter 2026 earnings call. I'm Sean Ricker, CFO of BigBear.ai, and I'm joined today by our CEO, Kevin McAleenan. Statements made in today's call that are not historical fact are considered forward-looking statements and are made pursuant to the safe harbor provisions of the federal securities laws. Actual results may differ materially from those projected in the forward-looking statements.
各位下午好,感謝大家今天參加我們 2026 年第二季財報電話會議。我是 BigBear.ai 財務長 Sean Ricker,今天與我一同出席的還有我們的執行長 Kevin McAleenan。今天會議中所作、非屬歷史事實的陳述,均屬前瞻性陳述,並依據聯邦證券法的安全港條款作出。實際結果可能與前瞻性陳述中所預期者有重大差異。
Please see today's press release and our SEC filings for a description of some of the factors that may cause actual results to differ materially from those in the forward-looking statements. We also have posted tables on our website today that we plan to address during the call to supplement our comments. These tables also include information regarding non-GAAP measures that may be used in today's call. Please access our website at www.bigbear.ai and click on the Investor Relations link to view and follow the charts.
請參閱今日新聞稿及我們向 SEC 提交的文件,以了解可能導致實際結果與前瞻性陳述所述結果出現重大差異的部分因素。我們今天也在網站上發布了表格,並計畫在會議中加以說明,以補充我們的評論。這些表格亦包含與今日會議可能使用之非 GAAP 衡量指標相關的資訊。請造訪我們的網站 www.bigbear.ai,點選「Investor Relations(投資人關係)」連結以查看並跟隨圖表。
Now I'll hand it over to Kevin.
現在我把時間交給 Kevin。
Kevin Mcaleenan - Chief Executive Officer, Director
Kevin Mcaleenan - Chief Executive Officer, Director
Good afternoon, and thank you, Sean. I'm pleased to report to our investors that BigBear.ai had a strong quarter. First, revenue for the second quarter of 2026 exceeded revenue for the same period in 2025 by 13%, and our gross margin improved by 781 basis points. We are on track to meet our top line guidance in the range of $135 million to $165 million.
各位下午好,謝謝你,Sean。我很高興向投資人報告,BigBear.ai 本季表現強勁。首先,2026 年第二季營收較 2025 年同期成長 13%,且我們的毛利率提升了 781 個基點。我們正按計畫達成營收指引,區間為 1.35 億至 1.65 億美元。
And as a leadership team, we are fully focused on following through, hitting our targets and continuing to earn investor trust. Second, we continue to win new contracts with both existing and new customers in the US and abroad. I'm happy to share that we won more than 20 new contracts with customers in quarter two with values ranging up to $5 million each.
作為領導團隊,我們全力以赴落實承諾、達成目標,並持續贏得投資人信任。第二,我們持續在美國及海外,向既有與新客戶贏得新合約。我很高興分享,我們在第二季與客戶簽下超過 20 份新合約,單筆合約金額最高可達 500 萬美元。
For confidentiality reasons, we can't share the details of many of them, but I'll touch on a few. This follows our announcement in Q1 of a longer-term contract with an intelligence customer with a total contract value of $53 million. Each win proves that customers want specialized AI for complex use cases and secure environments that very few companies have the operational insight to deliver. Customers see that we are focused on the right problems and have the capability to help solve them.
基於保密原因,我們無法分享其中許多合約的細節,但我會提及幾個例子。這也延續了我們在第一季所宣布、與某情資客戶簽訂的一份較長期合約,該合約總價值為 5,300 萬美元。每一次得標都證明,客戶需要用於複雜使用情境與安全環境的專業化 AI,而具備營運洞察、能交付此類能力的公司屈指可數。客戶看見我們聚焦於正確的問題,並具備協助解決問題的能力。
As a result, our backlog has grown by 9% since year-end, and the pipeline continues to be healthy. And third, thanks to a clear strategy and focused use of our development resources, we have deployed product enhancements and capabilities that are fundamentally changing operators' ability to drive mission performance across a range of use cases from cargo inspection, to generative AI, to drone swarm orchestration.
因此,自年末以來,我們的在手訂單(backlog)成長 9%,而商機管線(pipeline)也持續維持健康。第三,得益於清晰的策略與對研發資源的聚焦運用,我們已部署產品強化與新能力,從貨物檢查、生成式 AI 到無人機蜂群協同指揮等多種使用情境,正從根本上改變操作人員推動任務績效的能力。
I'll describe some of these examples shortly. These three facts that BigBear.ai is on track to forecast, realizing strong customer wins in multiple categories, and generating traction for product innovation, strengthen my confidence that we are building momentum as a company. We have laid excellent financial foundations, focused our talent on a disciplined strategy, and global market trends are in our favor.
我稍後會說明其中一些例子。BigBear.ai 正按計畫達成預測、在多個類別取得強勁客戶得標,並在產品創新上形成動能,這三項事實強化了我對公司正在累積成長動能的信心。我們已奠定優異的財務基礎,將人才聚焦於嚴謹的策略,而全球市場趨勢也對我們有利。
If I step back and consider what we thought would be true now when we began resetting BigBear in early 2025, our key underlying assumptions are playing out. Governments and business leaders globally are confronted by great power competition and rolling regional conflicts. BigBear.ai was prepared for this shift, thanks to our long-standing work for national security customers and our development of technologies targeted to meet new operational needs.
如果我退一步回顧,思考我們在 2025 年初開始重整 BigBear 時所預期、如今應該成真的情況,我們的關鍵底層假設正在逐步實現。全球各國政府與企業領袖正面臨大國競爭與持續發生的區域衝突。BigBear.ai 因長期服務國家安全客戶,並開發針對新作戰需求的技術,已為這樣的轉變做好準備。
This is the core of our expertise and customer trust. Our roots are deep in mission environments, supporting both domestic and international security. The trend line for increased investment in and focus on enhanced defense and security technologies is set, both in the US and among our allies. AI and alternative technologies are fundamental to the reimagined global battlefield and emerging military strategy. And commercial entities, especially those engaged in securing and facilitating trade and travel, urgently need advanced applied AI technologies, particularly for supply chain security.
這正是我們專業能力與客戶信任的核心。我們深耕任務環境,支援國內與國際安全需求。無論在美國或盟友國家,對強化國防與安全技術的投資與關注提升之趨勢已然確立。AI 與替代性技術是重新想像的全球戰場與新興軍事戰略的基礎。而商業機構,特別是致力於保障並促進貿易與旅行者,迫切需要先進的應用型 AI 技術,尤其是在供應鏈安全方面。
Domestic threat landscapes are far more dynamic as well. Counterterrorism professionals must sprint to stay ahead of threat actors who have access to powerful technologies at low cost. These asymmetric threats make BigBear.ai's technologies and experience ever more relevant to our core customers and of interest to new customers who would not previously have considered security as their priority. Critical infrastructure must be protected.
國內威脅態勢也更加動態。反恐專業人員必須全速前進,才能領先那些以低成本取得強大技術的威脅行為者。這些非對稱威脅使 BigBear.ai 的技術與經驗對核心客戶而言更具關聯性,也吸引了過去不會將安全視為優先事項的新客戶。關鍵基礎設施必須受到保護。
We predicted many of these changes when I led the Department of Homeland Security. Many of the threats we face now, including drone swarms and the potential for autonomous attacks on public events, as well as AI-enabled cyber threats were detailed in the DHS 2019 strategic framework for countering terrorism and targeted violence, and have become very real today.
我在領導國土安全部(DHS)期間就預測到其中許多變化。我們如今面臨的許多威脅,包括無人機蜂群與在公共活動中發生自主攻擊的可能性,以及 AI 強化的網路威脅,都曾在 DHS 2019 年《反制恐怖主義與針對性暴力戰略框架》中被詳細闡述,而如今已變得非常真實。
Additionally, the global travel and tourism industry is projected to reach an all-time high in 2026, accounting for 9.9% of global GDP, which represents 3.2% growth. In global trade, structural changes have led to increased investment in technology by companies and countries who must adapt to far more complex regulatory regimes.
此外,全球旅遊與觀光產業預計在 2026 年達到歷史新高,佔全球 GDP 的 9.9%,相當於 3.2% 的成長。在全球貿易方面,結構性變化促使企業與國家加大對科技的投資,因為他們必須適應更為複雜的監管體系。
Compounding shocks and the challenges of capacity and security have all driven broad reprioritization towards solving long-standing border management challenges with new technologies, including redesigning immigration systems. These issues sit at the heart of political agendas in both the developed and developing world. BigBear.ai builds and deploys technology to address each of these macro themes, which often converge.
多重衝擊疊加,以及產能與安全方面的挑戰,推動各界廣泛重新排序優先事項,轉而以新技術解決長期存在的邊境管理難題,包括重新設計移民系統。這些議題位於已開發與開發中國家政治議程的核心。BigBear.ai 建置並部署技術以因應上述每一項宏觀主題,而這些主題往往相互交會。
Capital in the private as well as public markets is increasingly invested in businesses with the experience to build, acquire and deploy defense and security technologies that will fundamentally change the world we live in, making nations stronger, safer and more prosperous. That's why as we enter the second half of 2026, we are increasing our pace to meet these opportunities.
無論在民間或公共市場,資本正日益投入於具備建置、收購並部署國防與安全技術經驗的企業;這些技術將從根本上改變我們所生活的世界,使國家更強大、更安全、更繁榮。因此,當我們進入 2026 年下半年,我們正加快步伐以把握這些機會。
My full focus is on two objectives: delivering top line revenue growth and accelerating our hunt for accretive M&A targets that offer catalytic technologies with clear potential for rapid deployment at scale. I'm confident that we can do this because as you will see from the examples that follow, we have now fully integrated both acquisitions we executed in the last six months, Ask Sage and CargoSeer.
我將全部精力聚焦於兩項目標:推動營收成長,以及加速尋找具增益性的併購(M&A)標的,這些標的能提供具催化作用的技術,並具備明確的潛力可快速且大規模部署。我有信心我們能做到,因為如同接下來的例子所示,我們已將過去六個月完成的兩項收購——Ask Sage 與 CargoSeer——全面整合。
These teams are now very much part of the day-to-day of the business and are performing well. While we will be deliberate about deploying our available cash reserves to acquire new targets, given the dynamic opportunities in very large addressable markets we are focused on, we believe this is time to be aggressive.
這些團隊如今已深度融入日常營運,且表現良好。儘管鑑於我們所聚焦、可觸及市場(TAM)極為龐大且機會動態多變,我們將審慎運用可用現金儲備來收購新標的,但我們相信現在是採取更積極行動的時機。
The BigBear.ai leadership team is ambitious and driving forward. My confidence in our ability to move fast rests on the simple fact that most companies today are searching to find their place in a world being redesigned by AI. BigBear.ai does not face that problem. Applied AI is at the very core of our value proposition and growth strategy and the operators we serve understand and are hungry to use AI where their use cases are real and urgent.
BigBear.ai 的領導團隊雄心勃勃,並持續推動向前。我對我們能夠快速行動的信心,建立在一個簡單事實之上:如今多數公司都在尋找自己在一個正被AI 重新設計的世界中的定位。BigBear.ai 不面臨這個問題。應用型 AI 正是我們價值主張與成長策略的核心,而我們所服務的營運單位也理解並渴望在其真實且迫切的使用情境中運用 AI。
To bring this to life, let me move now to three operational use cases for our technology. Each represents a recent enhancement and is underpinned by an example of the business that we have won with an active customer. The first is illustrated by a significant win in the last week for CargoSeer in El Salvador. We signed a five-year commercial deployment agreement with a strategic regional customs expert following a successful 12-month pilot.
為了讓這點更具體,我現在轉到我們技術的三個營運使用案例。每一個都代表近期的產品強化,並以我們與一位活躍客戶所贏得的業務案例作為支撐。第一個案例,來自我們上週在薩爾瓦多為 CargoSeer 拿下的一項重大勝利。在成功完成 12 個月試點後,我們與一位具策略性的區域海關專家簽署了為期五年的商業部署協議。
This is the first deployment of CargoSeer in Central America, bringing cargo X-ray imagery, import and export documents, and structured trade data into a single workflow. It helps El Salvador customs officials detect contraband, collect the correct duties and keep legitimate trade moving. Let's now look at the general application for CargoSeer.
這是 CargoSeer 在中美洲的首次部署,將貨物 X 光影像、進出口文件,以及結構化貿易資料整合到單一工作流程中。它協助薩爾瓦多海關官員偵測走私違禁品、徵收正確關稅,並讓合法貿易持續順暢運行。接下來我們看看 CargoSeer 的一般應用。
The operational challenges facing customs officers all over the world, including the increasing tension between growing trade volumes and increasing smuggling threats are substantial. The customs officer inspecting a shipment must evaluate rapidly assessing an X-ray image, a customs declaration and other trade data to judge whether the cargo is legitimate, often with a little more than a pen, a clipboard, and pages of manifest data.
全球各地海關人員所面臨的營運挑戰相當巨大,其中包括貿易量持續成長與走私威脅升高之間日益緊張的矛盾。海關人員在檢查一批貨物時,必須快速評估 X 光影像、報關申報資料及其他貿易數據,以判斷貨物是否合法;而他們往往只有一支筆、一個夾板,以及一頁頁的貨單資料可用。
An anomaly may signal a concealed contraband, misclassified commodity, or a trafficked human being, while legitimate shipments need to keep moving. The stakes are high and error can cost the country significant revenue, put lives at risk through drug or human trafficking or choke legitimate trade. That is the visibility, speed and volume problem that CargoSeer solves.
異常可能意味著藏匿的違禁品、錯誤分類的商品,或被販運的人口;同時,合法貨物仍需要持續流通。風險極高,任何錯誤都可能使國家損失可觀稅收、因毒品或人口販運而危及生命,或阻塞合法貿易。這正是 CargoSeer 所解決的可視性、速度與處理量問題。
Just as the doctor reads an X-ray alongside patient history to diagnose, CargoSeer analyzes the cargo scan alongside the entire data package powered by AI and driven by computer vision. The automated image analysis tool compares the image against the declaration and trade data and flag anomalies with the speed and explainability the officer needs to complete an inspection.
就像醫師會結合病史來判讀 X 光以做出診斷一樣,CargoSeer 以 AI 驅動並由電腦視覺支撐,將貨物掃描影像與完整資料包一併分析。自動化影像分析工具會將影像與申報資料及貿易數據進行比對,並以官員完成查驗所需的速度與可解釋性標示異常。
Why is this so valuable and why does it scale? Global merchandise trade runs roughly $26 trillion per year. As a former Port Director early in my career at LAX, I can attest that as a major operational challenge to balance protecting national security, collecting revenue and keeping legitimate trade moving with downward pressure on costs and staff. This is true in every port of entry and border crossing in the world, thousands of them, each facing overwhelming volume with limited staff and resources.
為什麼這如此有價值,而且為什麼能夠規模化?全球商品貿易每年約為 26 兆美元。我在職涯早期曾任洛杉磯國際機場(LAX)的港口主任,我可以證明:在成本與人力編制承受下行壓力的同時,必須在維護國家安全、徵收稅收與確保合法貿易順暢之間取得平衡,這是一項重大的營運挑戰。這在全球每一個入境口岸與邊境關卡都成立——成千上萬個據點,每一處都在有限人力與資源下承受龐大流量。
CargoSeer is configured to each country's own data and the officer's inspection workflow and its models can be updated as trade patterns and risks change. The technologies we are deploying in El Salvador are garnering interest from customs agencies around the world seeking to refine their operating models. The second product enhancement and use case is generative AI in disconnected environments.
CargoSeer 可依各國自身資料與官員的查驗工作流程進行配置,其模型也能隨著貿易型態與風險變化而更新。我們在薩爾瓦多部署的技術正吸引全球各地海關機關的興趣,他們希望藉此精進其營運模式。第二項產品強化與使用案例,是在斷線環境中的生成式 AI。
Last quarter, I shared that NASA is now a customer of our generative AI platform, Ask Sage. One of the wins we're particularly proud of in Q2 is with Naval Air Systems Command, NAVAIR, which provides full life cycle support for all aircraft, weapons and airborne systems used by the US. Navy and Marine Corps. I cannot detail the specific application of our technology, but its rapid software deployment and agentic coding capabilities are part of the use case.
上一季我曾分享,NASA 現在已成為我們生成式 AI 平台 Ask Sage 的客戶。我們在第二季特別引以為傲的一項得標,是與海軍航空系統司令部(Naval Air Systems Command,NAVAIR)合作;該單位為美國海軍與海軍陸戰隊所使用的所有飛機、武器與空載系統提供全生命週期支援。我無法詳述我們技術的具體應用,但其快速軟體部署與代理式(agentic)程式碼生成能力是該使用案例的一部分。
For customers who use our core cloud platform, two weeks ago, we expanded our offering for the entire Department of War and customers in other departments, including new air gap and local hardware for deployments up to the top secret level. This is a big step forward for the operator.
對於使用我們核心雲端平台的客戶,兩週前我們擴大了對整個戰爭部(Department of War)以及其他部門客戶的供應範圍,新增了用於最高機密等級部署的隔離網路(air gap)與本地硬體。這對第一線使用者而言是向前邁出的一大步。
Let me explain. Imagine an intelligence analyst in a secure facility comparing reports, searching mission data and preparing an assessment for a commander. The information cannot leave that environment and sometimes Internet connection is removed by design. The analysts still needs the latest AI tools to organize and evaluate data, but generative AI models are typically accessed through the cloud.
我來解釋一下。想像一位情報分析師在安全設施中比對報告、搜尋任務資料,並為指揮官準備評估。資訊不能離開該環境,而且有時網際網路連線會被刻意移除。分析師仍需要最新的 AI 工具來整理與評估資料,但生成式 AI 模型通常是透過雲端存取。
When an external network is unavailable or prohibited, that access disappears. Our new local device announced two weeks ago changes that. About the size of a hardcover book and highly portable, it brings the core capabilities of BigBear.ai's platform into locally connected or fully air-gapped environments that are physically isolated from all unsecured networks.
當外部網路不可用或被禁止時,這種存取就消失了。我們兩週前發布的新型本地裝置改變了這一點。它大約一本精裝書大小且高度便攜,能將 BigBear.ai 平台的核心能力帶入本地連線或完全隔離網路(air-gapped)的環境;這些環境在物理上與所有不安全網路完全隔離。
It is model-agnostic and multimodal. The analysts can use a range of frontier AI models and process text, images, video and more. And when a connection is available, teams draw on the broad catalog of cloud-hosted models. When air gapped, models hosted locally keep the platform running with no external network. This use case scales because our military and commercial customers can't afford service interruptions from patchy connectivity or security compromises.
它與模型無關(model-agnostic),並支援多模態。分析師可使用多種前沿 AI 模型,處理文字、影像、影片等。而當連線可用時,團隊可取用雲端託管模型的廣泛目錄。在隔離網路狀態下,本地託管的模型可在無外部網路的情況下維持平台運作。此使用案例之所以能規模化,是因為我們的軍方與商業客戶無法承受因連線不穩或安全受損而造成的服務中斷。
The operator uses the same BigBear generative AI platform across connected and disconnected environments instead of adopting a separate platform for each condition. Teams retain control over where their models run and the device can reach missions that cloud-only AI cannot serve. This expansion addresses two customer pain points, access and predictability, and directly supports the Department of War's AI acceleration strategy. We've also added a new delivery option to help customers manage the cost.
第一線使用者可在連線與斷線環境中使用同一套 BigBear 生成式 AI 平台,而不必針對不同條件各自採用不同平台。團隊可保有模型運行位置的控制權,而該裝置也能觸及僅靠雲端 AI 無法服務的任務。這項擴充解決了客戶兩大痛點:存取性與可預測性,並直接支援戰爭部的 AI 加速策略。我們也新增了一種交付選項,協助客戶管理成本。
Customers can buy tokens through BigBear or bring their own models and tokens if they have preferential pricing from the frontier labs, and they can also white-label our platform. All of these features are intended to offer maximum flexibility to our customers. The third product enhancement and use case centers on offensive and defensive drone warfare.
客戶可以透過 BigBear 購買 tokens,或在其從前沿實驗室取得更優惠價格時自備模型與 tokens,也可以將我們的平台進行白牌(white-label)化。所有這些功能都是為了向客戶提供最大的彈性。第三項產品強化與使用案例,聚焦於攻防無人機作戰。
Recently, in a series of military exercises, our ConductorOS software enabled a single soldier in the field to command multiple drones from multiple vendors running different models to clear an objective against notional enemy forces in a simulated degraded communications environment. Let's look more deeply at that use case. Today, a commander overseeing a mixed fleet of drones, sensors and mission software from different companies must operate systems that aren't often connected and can't share data, manually moving information between them to make it relevant while piecing together a changing mission across separate interfaces.
近期在一系列軍事演習中,我們的 ConductorOS 軟體讓一名在野外的士兵能夠指揮來自多家供應商、運行不同模型的多架無人機,在模擬的通訊受損環境中對抗假想敵軍並清除目標。讓我們更深入看看這個使用案例。如今,一位指揮官要監督由不同公司提供的混合無人機機隊、感測器與任務軟體,往往必須操作彼此不常連線、也無法共享資料的系統;他們需要在不同系統間手動搬移資訊,使其變得相關,同時在分離的介面中拼湊不斷變化的任務全貌。
The result is that massive data volume becomes a serious cognitive burden. ConductorOS gives those systems a common way to connect and coordinate, like a conductor leading an orchestra that unites the mixed fleet of hardware and software through an open vendor-neutral architecture that lets sensors and manned and autonomous platforms from different vendors share data, models, and inferences.
結果是,龐大的資料量成為嚴重的認知負擔。ConductorOS 為這些系統提供一種共同的連接與協同方式,就像指揮家帶領樂團一樣,透過開放且供應商中立的架構,將混合的硬體與軟體機隊整合起來,使來自不同供應商的感測器、有人與自主平台能共享資料、模型與推論結果。
Its reach extends beyond drone fleets across sensor networks, a wide array of autonomous systems, including air defense missiles and communications interfaces, but it is particularly powerful for drones, which have rapidly become integral to the modern battle space and are not yet operating autonomously across multiple manufacturers at scale. The case for scaling is clear.
其觸及範圍不僅涵蓋跨越感測器網路的無人機機隊,還包括各式各樣的自主系統(含防空飛彈與通訊介面),但對無人機而言尤其強大;無人機已迅速成為現代戰場不可或缺的一環,且目前尚未能在多家製造商之間以規模化方式實現跨廠商的自主運作。擴大規模的理由十分明確。
Soon, one operator will control large multi-vendor fleets rather than needing one operator per drone. In Ukraine, drone manufacturers have grown from seven private companies at the start of Russia's full-scale invasion to nearly 500 today, producing millions of units a year. In the 2027 defense budget request, the United States government called for more than $70 billion for drones and counter drone systems.
不久之後,一名操作員將能控制大型、多供應商的機隊,而不再需要一架無人機配一名操作員。在烏克蘭,無人機製造商已從俄羅斯全面入侵之初的七家民營公司,成長到如今近500家,每年生產數百萬台。在2027年國防預算請求中,美國政府要求為無人機與反無人機系統撥款超過700億美元。
It's the largest ever investment in the technology. Part of this is a $54.6 billion request for an autonomous warfare unit. Today, there are more than 400 drone companies in the US, and BigBear.ai is positioned to help them become operationally viable at scale. One further example beyond the drone application at the US Army's Jailbreak event in Fort Carson in May, ConductorOS integrated a sensor and missile system that had never been able to communicate, showing the sensor could pass location data on an incoming target directly to the missile system for a kinetic strike.
這是對該技術史上最大的一筆投資。其中一部分是為自主作戰單位提出的546億美元請求。目前美國有超過400家無人機公司,而BigBear.ai已就位,協助它們以規模化方式達到可投入作戰的營運能力。再舉一個超越無人機應用的例子:在5月於卡森堡舉行的美國陸軍Jailbreak活動中,ConductorOS整合了一套過去從未能互通的感測器與飛彈系統,展示感測器可將來襲目標的位置資料直接傳送至飛彈系統以進行動能打擊。
ConductorOS earned jailbroken status, meaning it had the approved interfaces and documentation needed for other systems to connect through it and operate as part of a broader interoperable network. This proves ConductorOS is on the leading edge of turning legacy system integration into a productized repeatable capability.
ConductorOS獲得「jailbroken」資格,意味著它具備經核准的介面與文件,使其他系統能透過它連接並作為更廣泛互通網路的一部分運作。這證明ConductorOS正站在將傳統系統整合轉化為可產品化、可重複交付能力的最前沿。
For operators deploying mixed fleets for multiple vendors, the need only grows as more autonomous platforms emerge. In each of these three use cases, we have advanced our technology, won new assignments and tackled operational problems that very few companies have the experience to understand, let alone address.
對於部署多供應商混合機隊的操作方而言,隨著更多自主平台出現,這項需求只會持續增加。在這三個使用案例中,我們推進了技術、贏得了新任務,並解決了極少數公司有經驗理解、更遑論處理的營運問題。
I'll now turn it over to Sean, to go to a deeper level on our financials.
接下來我把時間交給Sean,請他更深入說明我們的財務狀況。
Sean Ricker - Acting Chief Financial Officer
Sean Ricker - Acting Chief Financial Officer
Thanks, Kevin. Let's jump into our operating results for the second quarter. Revenue for the second quarter of 2026 was $36.7 million, an increase of 13% year-over-year as a result of performance from Gen AI platforms and products. Gross margin was 32.8% in the second quarter of 2026, an increase of 781 basis points as compared to the second quarter of 2025.
謝謝你,Kevin。我們直接進入2026年第二季的營運成果。2026年第二季營收為3,670萬美元,年增13%,主要來自生成式AI平台與產品的表現。2026年第二季毛利率為32.8%,較2025年第二季提高781個基點。
The expansion in gross margins was driven by a higher mix of revenue from Gen AI platforms and products versus the comparable period. Our net loss for the second quarter of 2026 was $25.7 million as compared to a net loss of $228.6 million in the comparable period. The decrease in the net loss was primarily related to noncash fair value changes in derivatives and a noncash goodwill impairment charge in the second quarter of 2025 that was not repeated in the second quarter of 2026.
毛利率擴張主要由於生成式AI平台與產品的營收占比較可比期間更高所帶動。2026年第二季淨損為2,570萬美元,相較於可比期間淨損2億2,860萬美元。淨損減少主要與衍生工具公允價值的非現金變動,以及2025年第二季認列但2026年第二季未再發生的非現金商譽減損費用有關。
Adjusted EBITDA for the second quarter of 2026 was negative $11.6 million versus negative $8.5 million in the comparable period. The decrease in adjusted EBITDA was primarily driven by increased investment in sales and go-to-market capabilities and in research and development, both of which were partially offset by expanded gross margins, as previously mentioned.
2026年第二季調整後EBITDA為負1,160萬美元,相較於可比期間為負850萬美元。調整後EBITDA下降主要由於我們加大對銷售與市場拓展能力,以及研發的投資;如前所述,部分影響由毛利率擴張所抵銷。
Turning to the balance sheet. Our balance sheet remains strong. We finished the second quarter with total cash and investments of $410 million, which we're using to invest in new capabilities and to seek accretive M&A that can accelerate our growth strategy. We ended the second quarter with $270 million in backlog, an increase of about $22 million from where we started the year.
接著看資產負債表。我們的資產負債表仍然強勁。第二季結束時,我們的現金與投資總額為4.10億美元,我們正用於投資新能力,並尋求具增益性的併購,以加速我們的成長策略。第二季末在手訂單(backlog)為2.70億美元,較年初增加約2,200萬美元。
As Kevin mentioned, we've seen contract wins across our portfolio, which reflects the dedicated focus and execution of our team in support of our customers and their missions. Before I turn it back to Kevin for closing remarks, I wanted to share an update relating to our execution rigor. In July, we passed a significant milestone. We officially achieved Cybersecurity Maturity Model Certification or CMMC Level 2, well ahead of schedule.
如Kevin所提,我們在整體產品組合上都取得合約勝利,反映團隊為支援客戶及其任務所投入的專注與執行力。在我把時間交回Kevin做結語之前,我想分享一則與我們執行嚴謹度相關的更新。7月,我們達成一個重要里程碑。我們正式取得網路安全成熟度模型認證(CMMC)第2級,且遠早於原定時程。
This is a rigorous US federal cybersecurity standard that proves we have the advanced security controls required to protect controlled and classified information. Achieving this means we have successfully verified our defenses against sophisticated cyber threats. The Department of War shared recently that it's reviewing the standard to lower the barrier to entry for smaller companies, but the threats that drove the CMMC Level 2 requirements are still here, and this certification is valuable to all our customers and also strengthens trust with international partners.
這是一項嚴格的美國聯邦網路安全標準,證明我們具備保護受控與機密資訊所需的進階安全控制。達成此項認證意味著我們已成功驗證自身防禦能力,可抵禦複雜的網路威脅。戰爭部近期分享其正在檢視該標準,以降低小型公司進入門檻,但促使CMMC第2級要求的威脅仍然存在;此認證對我們所有客戶都很有價值,也強化了與國際夥伴之間的信任。
The security practices it validates demonstrate a gold standard commitment to data privacy and infrastructure security. This is just one example of how we are, through our focus on executional rigor, maturing BigBear for future growth and expansion.
其所驗證的安全實務展現了對資料隱私與基礎設施安全的黃金標準承諾。這只是其中一個例子,說明我們如何透過對執行嚴謹度的專注,讓BigBear為未來的成長與擴張做好成熟準備。
Now I'll turn it back to Kevin.
現在我把時間交回Kevin。
Kevin Mcaleenan - Chief Executive Officer, Director
Kevin Mcaleenan - Chief Executive Officer, Director
Thank you, Sean. For the benefit of our new investors and as a reminder to loyal shareholders, I'll underscore that BigBear.ai is investing in our strengths as a specialized defense and security technology company, delivering mission-ready AI. We are focused on two core markets that are growing: national security, and trade and travel, as we've outlined; and we bring together three capabilities that make us different: deep mission understanding, an expert command of applied AI, and the unique combination of scale and agility vital for adapting and delivering as mission needs shift.
謝謝你,Sean。為了讓新投資人更清楚,也提醒長期股東,我要強調:BigBear.ai正以專精的國防與安全科技公司定位,投資於我們的強項,提供可立即投入任務的AI。如我們所述,我們聚焦於兩個正在成長的核心市場:國家安全,以及貿易與旅行;同時我們整合三項使我們與眾不同的能力:對任務的深度理解、對應用型AI的專家級掌握,以及在任務需求變動時能夠調整與交付所必需的規模與敏捷性的獨特結合。
Everything you heard today is evidence of that strategy in action. The accelerated program I talked about in Q1 is working. We have a realigned structure, clear growth plans and the resources we need to grow BigBear.ai into a powerhouse. To close, I believe that as attention moves away from legacy technology to the AI capabilities where implementation is vital, not a promise for the future, BigBear is positioned to win.
你們今天聽到的一切,都是這項策略正在落地的證據。我在第一季提到的加速計畫正在奏效。我們已重新調整組織架構,具備清晰的成長計畫,以及把BigBear.ai打造為強大企業所需的資源。最後,我相信,隨著市場注意力從傳統技術轉向AI能力(此時落地實施至關重要,而非對未來的承諾),BigBear已具備勝出的有利位置。
We operate equally well as a new defense and security technology prime or as a partner to the traditional primes, because we have deep customer trust and the size to deliver complex solutions or provide our targeted products and expertise within a larger contract.
我們既能作為新型國防與安全科技的主承包商(prime)同樣出色,也能作為傳統主承包商的合作夥伴,因為我們擁有深厚的客戶信任,以及交付複雜解決方案的規模;或是在更大型合約中提供我們具針對性的產品與專業能力。
That flexibility puts us in a highly advantageous position. During what will be an extended transitional period for big complex legacy businesses who need to pivot, BigBear is the ideal partner as we have already proven.
這種彈性使我們處於極具優勢的位置。在大型、複雜且需要轉型的傳統企業將面臨一段較長的過渡期之際,BigBear是理想的合作夥伴,而我們也已經證明了這一點。
I look forward to updating our shareholders on our next quarterly earnings call in November. Thank you for your interest in BigBear.ai.
我期待在11月的下一次季度財報電話會議上向股東更新進展。感謝各位對BigBear.ai的關注。
Operator
Operator
Thank you. This does conclude today's conference. We thank you for your participation.
謝謝。今天的會議到此結束。感謝各位的參與。