A SaaS product built on RAG, serving multiple customers from one system, has one non-negotiable requirement: customer A must never retrieve customer B's documents. A metadata filter alone is a fragile way to guarantee this — one bug in the filter logic and the guarantee is gone.
Stronger isolation options
- Separate indexes or namespaces per tenant, where the vector store itself enforces the boundary structurally, not just via a query filter.
- If sharing an index, enforce the tenant filter at the database or API layer, never trusting application code alone to always include it correctly.
- Test isolation explicitly: a test that logs in as tenant A and tries to retrieve tenant B's known documents should be part of your test suite, not an afterthought.
Why this deserves more caution than most RAG advice implies
A retrieval bug that leaks one wrong chunk into an answer is embarrassing. A retrieval bug that leaks one customer's confidential documents to another is a genuine incident. Design the isolation boundary as if a filter bug is inevitable, not unlikely.
GenAI engineer and AI product builder with 2+ years shipping production-grade AI systems: RAG pipelines, fine-tuned LLMs, hybrid retrieval and multi-modal apps across vision, speech and OCR, architected end to end from ingestion to deployment. Leads engineering for SCULT INDIA's 14-member team, founded the 500+ member TechVerse Enclave community and has mentored 200+ students. Three hackathon first prizes: Changethon 2025 (IIT Roorkee), Product Genesis at Vividhotsava 2025 (CSJMU Kanpur) and BYTEBATTLE (MeetKats).
Open to GenAI roles, hackathon judging, mentorship sessions and guest talks at colleges. Email pranjulrathour41@gmail.com.
