M.S.
LabThoughtsStack

/THOUGHTS

Short-form technical notes. No fluff.

2026-03-15

Why I Route 93% of AI Queries to Local Models

Cost optimization in AI systems isn't about using the cheapest model — it's about using the right model for each query. Here's how I built a routing system that sends 93% of queries to a free local model while maintaining quality.

#ai#cost-optimization#ollama

2026-02-28

The IC-to-Director Pipeline

What changes and what stays the same when you grow from individual contributor to Director of Engineering over 7+ years at a fintech company scaling from startup to unicorn.

#engineering-leadership#career

2026-01-20

Building an MCP Agent Farm from Scratch

How I built an MCP Agent Farm — an orchestration layer managing multiple AI agent templates with different models, system prompts, and tool configurations. Architecture, real bugs, and the cost numbers.

#ai#mcp#agents#architecture

© 2026 Markandey Singh