Cognition

Cognition is the first customer running NVIDIA Vera Rubin, powered by CoreWeave! Hopper in 2022. Blackwell in 2024. Now Vera Rubin: On SWE-2 inference, the new chips deliver about 4.8x more token throughput than GB200 at the same decode speed. More compute, better agents!

We’re thrilled to welcome Chris Degnan as our Chief Revenue Officer to help us scale globally and build the future of software. Chris joined Snowflake as its first sales hire. Over the next decade, he scaled alongside the company and grew into Snowflake’s CRO, driving over $3B in ARR and building one of the most iconic sales organizations in tech. He’s a world-class leader that builds winning teams, executes with relentless rigor, and focuses on what’s right for customers. We’re also excited to partner with Chad Peets and the RPT team to build Cognition’s GTM team. Chad was Chris’ co-pilot in building Snowflake’s world-class sales team. Welcome to Cognition, Chris.

With 12 engineers supporting 50+ attorneys, Crosby put Devin on first response. Devin investigates 20–40 bugs and alerts a day, cut Sentry noise by 50%, and kept on-call to one engineer as the legal team grew 5x.

Devin for MongoDB Modernizations is here. We’re bringing Devin into MongoDB’s Application Modernization Platform (AMP) to help teams move legacy applications to MongoDB Atlas. Modernization takes more than moving data. Engineers also have to change the application code that depends on the old database. Devin takes on that code work while AMP moves and validates the data. Engineers choose the target data model and cutover plan. Together, Devin and AMP can help customers reach production faster, so engineers can spend less time keeping legacy systems alive and more time building great software and solving hard problems. Learn more: https://lnkd.in/gEuCPEhn

Marlon Brandão de Sousa, a veteran engineer at EBANX, has a high bar for autonomous software engineering. He's blind. Before Devin, he had to listen to a screen reader describe errors so he could fix them. Now, Devin tests its work for him on its own computer and returns a merge-ready PR. His first major test: an integration planned for six weeks, completed in ~12 hours. Today, 92% of merged PRs in EBANX’s core payments platform are authored by Devin.

Devin now completes the same work with ~30-40% fewer ACUs Today we're rolling out updates that make every Devin session go further: → 30–40% cheaper in Fusion and Normal → 15–20% cheaper in Ultra → Up to 70% cheaper in Devin Review We didn't trade intelligence for savings. Performance held steady or improved across every mode, and Devin Fusion now ranks first on FrontierCode 1.1 Extended with a score of 68.8, at just $0.60 per task on average. Two things made this possible: 1. Model independence. Devin isn't locked into a single model. We choose the best fit for each part of the work, drawing on our own SWE-2, Opus 5.5, the GPT-6 family, and more. Fusion pairs a capable lead model with a cost-efficient sidekick. Devin Review picks the right models for analyzing diffs, finding bugs, and categorizing changes. 2. A more efficient harness. Newer models can plan more work per turn, so Devin now batches commands into single calls, runs independent actions in parallel, and makes the most of prompt caching. It reuses context instead of reprocessing it from scratch every turn. The result is that your Devin usage stretches further, so you can take on your most ambitious work. Read the full breakdown: https://lnkd.in/g6fpsfDs

Devin Mobile is now available in beta. Join the waitlist: https://devin.ai/ios

Cognition has crossed $1B in annualized revenue run rate. Every company today is a software company. We started Cognition in January 2024 to help them build more. Less than three years later, Devin works alongside engineering teams at GE Aerospace, Rivian, Exa, and many more. This milestone belongs to our customers. Here are a few of their stories.

Devin now works across Microsoft Teams and Microsoft 365. More than 90% of the Fortune 500 use Teams for everything from bug reports to product specs. Starting today, Devin can pick up work where those conversations already happen and share its results in the same place. In Microsoft Teams: → DM Devin or add it to a group chat to start a session → Set up Automations so Devin starts on its own when a bug report, alert, or request comes into a channel → Answer questions, approve actions, and sign in to MCP servers with one-click cards → Devin can @-mention coworkers, send file attachments, and mute itself when it's done With our new first-party Microsoft 365 integration, you can connect Devin to Outlook mail and contacts, Calendar, OneDrive and SharePoint, Microsoft To Do, Teams chats, and your Entra ID directory. Microsoft 365 uses delegated, per-user permissions only. Devin acts as the signed-in user and can reach only what that user can. Available today for all Devin customers.

Autonomous AI engineering is moving from pilots into production. Cognizant announced new work with Odyssey Logistics powered by Devin, helping modernize core legacy workflows into cloud-native software for OdysseyONE, Odyssey’s transportation management platform. This is the enterprise model we believe in: AI agents doing real engineering work inside governed delivery, with experienced teams reviewing, directing, and owning every change. We’re excited to continue the work we’re doing with Odyssey and Cognizant. See what customers are shipping with Devin: https://devin.ai/customers