Designs, builds and runs production features powered by large language models: retrieval-augmented assistants, agents that use tools, and AI capabilities inside SaaS products, built on OpenAI, Anthropic and open-source models. Owns the outcome that AI features are accurate, safe, fast and cheap enough to ship to real users and stay in production, proven by evaluation rather than demos.
Now hiring
PythonTypeScript / Node.jsOpenAI APIAnthropic API (Claude)Azure OpenAIAWS BedrockGoogle Vertex AI / Gemini APILangChain / LangGraphLlamaIndexVercel AI SDK
How it works
Five steps from CV to the bench.
1ApplyUpload your CV. We build your profile from it and you check it. About 10 minutes.
2ScreenA recruiter reviews it within 2 business days and books a 15-minute Teams call.
3TestShort, practical skills tests for your role. You get kind, specific feedback.
4Level
By applying you agree we can process your application as described in our privacy policy.
Your level (Intern to Specialist) sets your pay. Same pay for every client.
5BenchJoin the bench and get matched to clients in Australia, the UK and the US.
What you get
Paid on time, set by levelYour pay comes from your level, never negotiated per client.
Equipment and securityManaged laptop and security setup for Seat roles.
Your details stay privateClients see your evidence card: first name, level, scores and work. Never your surname, contact details or CV.
CareersSign in