Senior AI/ML Engineer

AurvikAI

Own AI components end-to-end for enterprise clients — retrieval design, model selection, deployment, and the iteration loop after launch. You'll work directly with founders and senior engineers, no proxy layers.

What you'll do

▪Design RAG pipelines on top of vector stores, hybrid search, and reranking — tuned to client domain data and evaluated against real user queries.
▪Choose, fine-tune, and benchmark open and closed LLMs. Build eval harnesses that catch regressions before users do.
▪Ship multi-agent workflows with tool use, planning, and guardrails for narrow enterprise problems.
▪Own production quality: latency, cost, observability, and safety. You don't hand the model off to ops — you keep it healthy in prod.

What we're looking for

▪5+ years building production software, with at least 2 shipping AI/ML systems that real users depend on. Research-only backgrounds aren't the right fit.
▪Deep Python and the modern AI stack — PyTorch or JAX, Hugging Face, LangChain or LlamaIndex (or strong opinions about why not), vector databases, eval frameworks.
▪Strong systems sense: you understand cost, latency, and failure modes, and know when smaller model + good retrieval beats bigger model + naive prompts.
▪Clear written communication. You can write a one-page design doc a non-AI engineer can follow.

How to apply

Email career@sdtcdigital.com with your CV (PDF), a short note about a production AI system you've shipped (what it did, what you owned, what you'd do differently), and links to anything we can read or run.