Consulting
Vector search & RAG consulting
Hands-on help from an engineer with 6+ years running search in production, including multi-region OpenSearch on AWS.
Who it’s for
Teams building RAG, vector search or agent retrieval on AWS, especially:
- Teams building RAG features and unsure their retrieval is good enough
- Products adding vector or hybrid search to an existing search stack
- Agent builders who need reliable retrieval tools
- Teams running OpenSearch on AWS who want it faster or cheaper
What we offer
- 01
Architecture review
A review of your retrieval stack, from ingestion and chunking to indexing, query flow and serving, with a written report of risks and the changes worth making first.
- 02
Retrieval quality evaluation
A repeatable way to measure whether your search or RAG pipeline returns the right context: labelled query sets, offline metrics and a baseline you can track release to release.
- 03
OpenSearch performance & cost tuning
Shard and index layout, k-NN and HNSW settings, instance sizing and query patterns, aimed at lower latency and a smaller bill on Amazon OpenSearch Service.
- 04
Embedding model migrations
Moving to a new embedding model without downtime or a quality drop: evaluation, dual indexing, backfills and a cut-over plan.
Engagements start at $2,000 USD
Tell us what you’re building and where it hurts. We’ll reply with whether we can help and a suggested scope.