← All jobs · Cognition

Special Projects Engineer - UK

Cognition ·
75
AI-Agency
B72 U80
📍 London, GB 🛠 AI tools welcome at work Mid 2+ yrs
DevinWindsurf
TL;DR

Special Projects Engineer at Cognition, an applied AI lab building software agents (Devin, Windsurf). Role combines engineering, product, and GTM—owning end-to-end customer implementations and translating pain points into product improvements.

Apply at Cognition →
share:
you'll be redirected to the company's career page

Job description

We are an applied AI lab building end-to-end software agents.

We're the makers of Devin, the first AI software engineer, and Windsurf, the AI-native IDE. Together, they represent our vision for collaborative AI teammates that enable engineers to focus on more interesting problems and empower teams to strive for more ambitious goals.

Our team is small and talent-dense. Among our founding team, we have world-class competitive programmers, former founders, and leaders from companies at the cutting edge of AI including Scale AI, Palantir, Cursor, Waymo, Tesla, Lunchclub, Modal, Google DeepMind, and Nuro.

Building Devin and Windsurf is just the first step—our hardest challenges still lie ahead. If you’re excited to solve some of the world’s biggest problems and build AI that can reason on real-world tasks, apply to join us.

About the Role

Equal parts Engineering, Product, and GTM – perfect for a former founder or aspiring one. You’ll translate real-world pain into product improvements, hands-on implementations, and measurable impact, by whatever means necessary.

Core Responsibilities:

What We Value:

What We Require:

Equal Opportunity

Cognition is an equal opportunity employer. We do not discriminate on the basis of race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability, veteran status, or any other protected characteristic under applicable law. We are committed to providing reasonable accommodations for candidates with disabilities throughout the hiring process - please let us know if you need any.

Apply at Cognition →

More open roles at Cognition

Cognition ⚡ AI-native · 🔄 synced 5h ago
Research, Post-Training
📍 San Francisco, US 🛠 AI tools welcome at work · Senior
Research scientist at Cognition building post-training methods for AI agents. Focus on training recipes, evaluation design, alignment techniques (RLHF, RLAIF), and scaling methodologies for Devin and Windsurf.
RLHFRLAIFPyTorchJAXdistributed training
93
AI-core
Cognition ⚡ AI-native · 🔄 synced 5h ago
Research, Mid-Training
📍 San Francisco, US 🛠 AI tools welcome at work · Mid
Research scientist at Cognition focused on mid-training optimization for large language models. Responsibilities include data mix design, synthetic data pipelines, annealing schedules, and context length extension to improve model reasoning and coding capabilities.
PythonPyTorch
91
AI-core
Cognition ⚡ AI-native · 🔄 synced 5h ago
Product Engineer
📍 San Francisco, US 💰 $260K–$300K 🛠 AI tools welcome at work · Mid
Product Engineer at Cognition building end-to-end software agents and AI-native developer tools. Own features across IDE, web, and CLI surfaces; work directly on Devin agent infrastructure and developer experience.
PythonTypeScriptReact
87
AI-core
Cognition ⚡ AI-native · 🔄 synced 5h ago
Research, Post-Training Data
📍 San Francisco, US 🛠 AI tools welcome at work · Senior
Research scientist at Cognition designing post-training data strategies for AI agents. Focus on human preference modeling, synthetic data generation, and evaluation design to improve model behavior and alignment.
PythonPyTorchTensorFlow
85
AI-core
Cognition ⚡ AI-native · 🔄 synced 5h ago
Research Engineer, Infrastructure
📍 San Francisco, US 🛠 AI tools welcome at work · Senior
Research Engineer, Infrastructure at Cognition building distributed training systems and experiment orchestration for large-scale AI agent development. Own GPU cluster infrastructure, data pipelines, and performance optimization across thousands of GPUs.
PythonC++PyTorchGPUKubernetesdistributed systems
83
AI-core