Real-World RAG Incidents: Lessons from a Production Rollout
KU
Kiril urbonas
April 4, 2025 • 2 min read•Updated last week•3 views
More from AI
Explore more articles in this category
Embeddings Drift Detection — When "Similar Enough" Stops Being Similar
Embedding indexes degrade silently. The signals that catch drift, how often to re-embed, and the operational patterns we built after one quiet quality regression.
7 min read
LLM Streaming UX — Backpressure, Cancellation, Partial Results
Streaming LLM responses is easy until the client disconnects, the model stalls, or the user cancels. The patterns that keep streaming responsive without leaking spend.
8 min read
AI Agent Tool Design — Boundaries and Confirmations
When LLMs can call tools that change real state, the design decisions that matter most are about what's gated, what's automatic, and what triggers a human checkpoint.
8 min read
