The work you’ll own.
Turn ambiguous business processes into dependable software. You will work across document understanding, workflow orchestration and evaluation, with direct ownership of the systems you ship.
What you’ll do.
- Map customer workflows and identify where AI adds value and where deterministic code is safer.
- Build retrieval and document-processing pipelines with source references, structured output validation and human approval.
- Connect business tools through authenticated APIs, webhooks, queues and retry-safe jobs.
- Create evaluation datasets, regression checks and operational dashboards before expanding a workflow.
What you’ll bring.
- Experience shipping Python or TypeScript services and working with SQL databases.
- Practical experience with an LLM API, structured outputs, retrieval and evaluation.
- Comfort with authentication, background work, rate limits and failure recovery.
- Clear written communication and a portfolio or repository that explains your decisions.
What good work looks like.
- A mapped workflow with an explicit success measure.
- An integration with meaningful automated checks and observable failure states.
- A documented handover another engineer can operate.
A few practical details.
This is a full-time role, working remotely within India. Compensation is discussed individually before an offer, based on the role, experience and agreed scope. We’ll discuss collaboration hours and availability during the conversation.
Apply by 13 November 2026. Include a CV, portfolio or relevant repository and explain your contribution to one project. Please leave out identity documents, payment information and confidential client material.
← All opportunities