Backend Engineer — E-commerce Semantic Search (4-Month Project)

This is a fixed-term project, not an open-ended role: four months, part-time, for a mid-size e-commerce client who wants their product search to actually understand what people are looking for — not just match keywords.

You'd own the RAG pipeline end-to-end: turning a product catalog into something searchable by meaning, not just SKU titles, and making sure it holds up once real customers start typing real, messy queries into it.

• Design and build a RAG-based semantic search pipeline for an e-commerce product catalog
• Handle chunking/indexing of product data (titles, descriptions, attributes) and retrieval tuning
• Integrate the search backend with the client's existing e-commerce stack
• Set up basic evaluation — knowing whether a change made search better or worse, not guessing
• Hand off a working, documented system by the end of the engagement — this project has a defined end date

• Real hands-on experience building RAG or semantic search systems — specifics, not "I used an embeddings API once"
• Strong Python
• Comfortable working independently on a fixed scope, without a long ramp-up — this is a 4-month engagement
• Available part-time for the full 4 months — we need continuity, not someone who disappears mid-project

• Prior e-commerce search experience — ranking, filters, the kind of edge cases a real product catalog throws at you
• Experience with vector databases at catalog scale (tens of thousands+ SKUs)

• Remote, part-time — work around your own schedule as long as the 4 months land on time
• A contained, well-scoped project — clear start, clear end, clear deliverable
• Real client, real production traffic, not an internal experiment