Home Work Engineering Get in touch

Notes from
production.

Technical articles on AI systems, architecture decisions, and engineering tradeoffs — written from real-world experience building production platforms.

Building Intelligent
Lead Matching Systems

A five-part series exploring the architecture and engineering decisions behind an AI-powered lead matching pipeline — from basic keyword matching to production-grade hybrid retrieval and LLM validation.

01
Building Semantic Lead Classification Using Embeddings and Elasticsearch
AI Systems · 10 min read · May 2026
Published
02
Why Semantic Search Alone Was Not Enough for Lead Matching
AI Systems · Architecture
Coming soon
03
Combining Embeddings and LLM Validation for Business Matching
AI Systems · Engineering Decisions
Coming soon
04
Challenges Validating LLM Outputs in Production Systems
Engineering Decisions · Architecture
Coming soon
05
Lessons Learned Building AI Pipelines with Rails and Ollama
Architecture · Engineering Decisions
Coming soon
🧠
AI Systems

Practical articles on building and operating AI-powered systems in production environments.

Embeddings LLMs Semantic Search Retrieval Pipelines Validation
⚙️
Architecture

System design, scalability patterns, and the structural decisions that make platforms reliable at scale.

Distributed Systems Workers Queues Scaling Observability
⚖️
Engineering Decisions

The tradeoffs, failures, and lessons that shape how systems get built and rebuilt over time.

Tradeoffs Why X Why Y Failed Lessons Learned