Cognition
Tech / AI / Software
SoftwareEngineer
Neural analysis suggests this role is
optimal for Mid+ candidates.
“Software Engineer at Cognition. Skills: Systems engineering depth, Python proficiency, applied AI, agent infrastructure, tool use, context management, multi-step planning, subagent orchestration, sandboxed code execution environments, editor intelligence, agent-in-the-loop workflows, real-time code understanding, developer experience, reliability and performance at scale. Build core agent infrastructure. Design and ship the systems that power Devin's long-horizon task execution: tool use, contex”
Industry & Context.
reason across thousands of lines of code; use tools reliably across ambiguous long-horizon tasks; hard problems with incomplete specs; learn fast from results; course-correct without needing a lot of direction; how agents fail; what it takes to make AI-powered systems behave reliably in the real world
What They're Looking For.
Must Have
Systems engineering depth, Product instinct, Comfort with ambiguity, Velocity without shortcuts, Curiosity about agents and AI, Python proficiency, Relevant industry experience, Degree from a top-tier university: BS, MS, or equivalent in Computer Science, Mathematics, Engineering, or a related technical discipline from a highly selective program
Nice to Have
Prior experience at a frontier AI lab, applied AI company, or developer tools company
What You'll Do.
Build core agent infrastructure
Design and ship the systems that power Devin's long-horizon task execution: tool use
subagent orchestration
and sandboxed code execution environments
Improve Windsurf as an AI-native IDE
Contribute to editor intelligence
agent-in-the-loop workflows
real-time code understanding
and the developer experience that makes Windsurf different from every other IDE
Close the loop between models and products
Work directly with researchers to translate new model capabilities into shipped products
Own reliability and performance at scale
Build systems that handle millions of agentic tasks with low latency
and the kind of correctness that developers depend on in production
Move the category forward
How You'll Work.
Team & Collaboration
Work directly with researchers to translate new model capabilities into shipped products; Your feedback shapes what gets prioritized in training
Full Job Description
WHO WE ARE Cognition is an applied AI lab building end-to-end software agents. We are behind Devin, the first AI software engineer, and Windsurf, an AI-native IDE. Our vision is AI that works alongside engineers as a genuine teammate, not a tool. We are a small, talent-dense team of competitive programmers, former founders, and researchers from Scale AI, Palantir, Cursor, Google DeepMind, and others. ROLE MISSION Software Engineers at Cognition are not feature builders. You will be working on some of the hardest open problems in applied AI: how do you build an agent that can reason across thousands of lines of code, spawn and coordinate subagents, use tools reliably across ambiguous long-horizon tasks, and do all of this in a way that a real engineer would trust? You will ship systems that go directly into Devin and Windsurf, two products that millions of developers use to write, debug, and ship code. This is a role for engineers who want to be close to the frontier, who can move fast without cutting corners, and who believe the next 5 years of software engineering will look fundamentally different from the last 5. WHAT YOU'LL ACCOMPLISH - Build core agent infrastructure: Design and ship the systems that power Devin's long-horizon task execution: tool use, context management, multi-step planning, subagent orchestration, and sandboxed code execution environments. - Improve Windsurf as an AI-native IDE: Contribute to editor intelligence, agent-in-the-loop workflows, real-time code understanding, and the developer experience that makes Windsurf different from every other IDE. - Close the loop between models and products: Work directly with researchers to translate new model capabilities into shipped features; your feedback shapes what gets prioritized in training. - Own reliability and performance at scale: Build systems that handle millions of agentic tasks with low latency, high reliability, and the kind of correctness that developers depend on in production. - Move
Applying for this Software Engineer role?
Most applicants get filtered before a human reads their resume. See if yours makes the cut.
How to Apply on Ashby
- Ashby is a fast modern ATS — most applications take under 3 minutes.
- The resume parser is strong; verify parsed experience dates and job titles.
- Custom screening questions are often scored algorithmically — answer completely.
- Location field affects geo-based screening; use your actual metro area.
ANONYMOUS · UNFILTERED
What do employees actually say about Cognition?
Real rants from real employees. Read before you apply.