BLOG

Writing

On LLM systems, backend engineering, databases, and Web3.

LLM8 minJun 2026

Building a multi-provider fallback chain that actually fails over

Claude → Gemini → OpenAI: how to build a fallback chain that handles rate limits, partial failures, and observability without turning into spaghetti.

RAG11 minJun 2026

RAG across 40+ languages: what broke and what fixed it

A LangChain pipeline with Gemini + OpenAI fallback, and the multilingual edge cases nobody warns you about.

Cost6 minJun 2026

Cutting LLM spend ~95% with 24h prompt caching

When your system prompt is large and stable, caching is the cheapest win on the board. Here is how to do it across providers.

Architecture9 minMay 2026

Designing RBAC with sub-ms access checks on MongoDB

Sparse indexes, role hierarchies, and keeping authorization off the hot path in a multi-owner CRM.

Databases7 minMay 2026

Zero-downtime migrations in 8 feature-flagged phases

How to move a live data model without the sales floor ever noticing. A walkthrough of the Meraki migration.

Solidity10 minApr 2026

Building a multi-sig wallet in Solidity, step by step

Threshold approvals, replay protection, and deploying to Polygon testnet. A complete walkthrough.