← All rolesOpen role · direct hire
AI Integration Engineer
Put language models where they earn their place — screening, classification, intelligence, drafting — grounded in our own data, with the evaluation discipline to keep them honest.
What you'll do
- Build agent features on the Anthropic SDK with structured, typed outputs.
- Engineer retrieval (Qdrant + Voyage) so outputs are grounded and cited, not generic.
- Treat prompts as code: versioned, tested, evaluated; defend against prompt injection.
- Keep cost, latency, and a human approval gate firmly in view.
What we're looking for
- Strong engineering in Go or TypeScript.
- Hands-on with LLM APIs and RAG / embeddings in something real.
- An evaluation mindset — you measure model output, you don’t trust vibes.
Nice to have
- Go
- Vector databases
- Safety / prompt-injection experience
- Data plumbing
What success looks like in 90 days
- Shipped a grounded, evaluated agent feature.
- Stood up a small eval harness others use.
- Cut a cost or quality problem with data, not guesswork.
Stack
GoTypeScriptAnthropic SDKQdrantVoyage
The process
01
Apply
CV + cover letter
02
Screen
we read every one
03
Assessment
a real, scoped task
04
Interview
the team you’d join
05
Founder talk
values & direction
06
Offer
terms discussed openly
07
Onboard
paired, shipping in week one
Apply for this role
You’re applying for AI Integration Engineer. Your documents upload securely to our storage and only you and our hiring team can see them — they are reviewed by a person, not discarded into an inbox.
⬆Drag & drop, or browsePDF or Word · up to 8 MB
▹ PDF or Word, up to 8 MB. Files upload directly to secure storage; we receive only the reference. Your data is handled per our Privacy Policy and never sold.