About The Role
Own the grounding layer behind our AI products. You will build retrieval that actually finds the right passage, and the evaluation that proves it, because an ungrounded model is a liability with a confident voice.
What You'll Do
Design chunking, embedding and hybrid retrieval strategies
Build and tune vector search over real, messy document sets
Implement citation and provenance so answers can be checked
Build golden-set evaluation for groundedness and correctness
Tune refusal behaviour so thin retrieval escalates instead of improvising
Optimise token cost and latency per query
What You'll Bring
2+ years backend or ML engineering
Solid Python and practical LLM API experience
Hands-on with vector databases and embedding models
Understanding of why hybrid search beats pure semantic search
Evaluation instincts: you measure before you claim
Nice to Have
pgvector, Pinecone or Weaviate in production
Reranking and query-rewriting experience
Multilingual retrieval experience
What You'd Build
These aren't hypothetical projects, they're live products you can try before your first interview.
The StackBinary MarTech Suite, The full product line, every system we ship, most with live demos.
Why Join StackBinary™?
Flexible working hours
Remote-friendly culture
Learning & development budget
High-ownership projects
Pragmatic engineering culture
Work with cutting-edge tech
Ready to Apply?
Join our team of builders who love shipping quality software.
We connect with shortlisted candidates through LinkedIn or our official email IDs.
Follow StackBinary on LinkedIn →
Questions about this role?
[email protected]
📌 LLM & RAG Engineer (Mumbai)
🏢 Stackbinary
📍 Mumbai