AI, LLMs, and applied ML.

Senior-engineer field notes on AI, LLMs, agents, and applied machine learning by Ercan Ermis.

All posts

66 posts

Pick a Vector Database by What Breaks First

Almost every RAG system should start on pgvector. The useful question is not which vector database is fastest, but which failure mode reaches you first.

Jul 29, 2026 12 min

One Door to the Models, Part 10: Observability, Governance, and the Numbers

Logging prompts makes the gateway a processor of whatever users typed. Part 10 closes the series on traces, production evals, governance, and the scorecard.

Jul 28, 2026 8 min

One Door to the Models, Part 9: Eval Gates and Model Retirement

An approval that times out marks the stage skipped, not failed. Part 9 makes the eval a real gate and turns a model retirement into a query, not a project.

Jul 27, 2026 8 min

One Door to the Models, Part 8: Orchestration on Top of the Gateway

An orchestration framework that retries and reroutes is competing with the gateway. Part 8 keeps the framework thin and puts tools under the same identity.

Jul 26, 2026 7 min

One Door to the Models, Part 7: Azure AI Search or a Vector Database

Azure AI Search bills capacity by SKU, not by vectors, and the vector index is capped by tier memory. Part 7 decides retrieval on limits, not features.

Jul 25, 2026 8 min

One Door to the Models, Part 6: Semantic Caching and Its Failure Modes

A semantic cache is a correctness surface, not just a cost lever. Part 6 tunes the threshold, isolates tenants, and handles the day the cache is gone.

Jul 24, 2026 8 min

One Door to the Models, Part 5: Identity, Quota, and Chargeback

You get five custom metric dimensions, and time series multiply. Part 5 builds tenant identity, quota, and a chargeback model that survives cardinality.

Jul 23, 2026 9 min

One Door to the Models, Part 4: Async Work Off the Request Path

Batch inference is a second door: a file goes in, a job runs, a file comes out, and no gateway policy is in the path. Part 4 keeps the accounting honest.

Jul 22, 2026 10 min

One Door to the Models, Part 3: The Provider Abstraction and Streaming

Set stream to true and token counting becomes estimation. Open a WebSocket and load balancing stops existing. Part 3 maps guarantees to transports.

Jul 21, 2026 11 min

Enterprise MCP Auth: Agents Finally Get Service Accounts

SEP-1046 gives unattended agents a real OAuth story on MCP: client credentials with private_key_jwt. What it fixes, what it leaves open, and how to roll it out.

Jul 21, 2026 9 min

More from Ercan

Two more sites, same author, different ground.