Building Multi-Tenant Agent KBs: Is the 3-Layer Architecture Overengineered?
Present-Entry8676 · reddit · 2026-08-11
A developer building a multi-tenant AI Agent orchestration system in Go discusses the architecture of Knowledge Base (KB) pipelines and seeks community advice.
Proposed 3-Layer Architecture
- Source: Where data lives (S3 buckets, crawled websites, Notion workspaces).
- Normalizer: Converts various formats (PDF, HTML, Word) into a consistent format (e.g., Markdown), so the rest of the pipeline doesn't care about the original.
- Index: Chunks normalized content and makes it searchable via vector indexes (pgvector) or full-text search (Postgres tsvector).
The Core Dilemma
While decoupling is elegant in theory, it forces users to make 3 decisions. This is overwhelming for average users who just want to 'upload a PDF and make the bot smart.'
Proposed Solution
The author considers hiding complexity behind presets (e.g., a 'Documents' preset wiring S3 + default normalizer + vector index), exposing granular controls only in an 'advanced mode.' The community discusses if this pluggability pays off, Markdown's limitations, and the trade-off between presets and control.
More from coding & agent
- Insight: Designing Agent Evals is Like Designing Take-Home Interviews — andrew_n_carr · 2026-08-11
- Indie Dev Experience: VPS-Hosted Tools Require Almost Zero Maintenance — JamesIvings · 2026-08-11
- MiniMax H3 All-in-One Creator Node for ComfyUI Released — Fine_Rhubarb3786 · 2026-08-11
- Open-source phone-harness lets Claude Code control your iPhone like a human — moeinteractive · 2026-08-11
- YC S26 Startup Async Launches Specialized AI Agents for SMB Operations — ycombinator · 2026-08-11
- New ComfyUI Node Enables Negative Prompts for Krea 2 Turbo at CFG 1 — 1wndrla17 · 2026-08-11