Day 5:Week 10, Day 5: Metadata & RBAC Security — Why Post-Filtering Fails (`Zero-Results Cliff`) vs Pre-Filtering
Battle-tested daily lesson note • 100% runnable code, interview Q&A, test suite & interactive quiz
Week 10, Day 5: Metadata & RBAC Security — Why Post-Filtering Fails (Zero-Results Cliff) vs Pre-Filtering
1. The Multi-Tenant & RBAC Security Requirement in Enterprise RAG
In any real company, a user from tenant_id="acme_corp" or role="junior_support" must NEVER see chunks belonging to tenant_id="globex" or role="executive_board".
Every vector query has both a Vector Similarity Search AND a Boolean Metadata / ACL Predicate:
WHERE tenant_id = 'acme_corp' AND allowed_role IN ('public', 'junior_support')
2. The "Zero-Results Cliff" of Naive Post-Filtering
What happens if you run a global HNSW search for top_k = 10 across the entire 1,000,000-chunk database first, and then filter those 10 results in Python (Post-Filtering)?
Suppose tenant_id = 'acme_corp' owns 0.5% (5,000 chunks) of the database:
- The expected number of
acme_corpchunks inside the global unfilteredtop_k = 10is $10 \times 0.005 = 0.05$! - 95% of the time, Post-Filtering returns
0results ([]) even thoughacme_corphas5,000matching documents!
Naive Post-Filtering (BROKEN on Selective Filters!):
[1,000,000 Total Vectors] ──► Global Top-10 ANN ──► Filter by tenant='acme' (0.5%) ──► 0 Results!
Pre-Filtering / Single-Stage Filtered ANN (PRODUCTION STANDARD):
[1,000,000 Total Vectors] ──► Inverted Bitmask for tenant='acme' (5,000 IDs) ──► Rank Allowed IDs ──► Full Top-10!
How Production Vector DBs (Qdrant, Milvus, pgvector) Execute Filtered Search
Maintain an Inverted Metadata Bitmask Index mapping each metadata key-value pair (tenant_id="acme_corp") to a roaring bitmap of valid integer vector IDs:
- If the filter selectivity is small (e.g.,
< 5,000matching IDs), directly score the bitmask-selected vectors via SIMD (Pre-Filtering)—which is actually faster than graph traversal and guarantees100%recall! - If the filter matches a larger subset, evaluate the bitmask predicate during
HNSWgraph neighbor expansion (Single-Stage Filtered HNSW).
In Simple English: Why Filtering Security Permissions AFTER Search Fails
Imagine a company with 10,000 HR Payroll PDFs (Admin-only) and 200 Engineering FAQs (Public):
- Post-Filtering Disaster ("Zero-Results Cliff"): Suppose an intern asks "vacation policy". Vector search grabs the Top 5 overall matches — all 5 turn out to be Admin HR PDFs! Then the security filter deletes all 5, handing the intern 0 results even though 200 public FAQs existed!
- Pre-Filtering Fix: First mask down to ONLY documents the user is allowed to see (
tenant_id + role_acl) using roaring bitmaps / B-Tree metadata indexes, THEN run vector search inside that authorized slice! Never leak unauthorized docs OR return empty screens!