AI that ships, not slides.
What we're learning building production AI — RAG, computer vision, LLM agents, and the stuff nobody puts in the demo.
RAG in Production: What Nobody Tells You About Chunking
We've built 6+ RAG systems. Fixed-size chunking breaks on tables, multi-column PDFs, and nested lists. Here's what actually works — and the evaluation pipeline you need before shipping.
We Replaced 80% of Manual QA with Computer Vision
A real deployment on a 24/7 manufacturing line. Edge inference, defect grading, and the accuracy numbers after 6 months.
Claude vs GPT-4 for Code Generation: Our Internal Benchmark
We ran both models on 200 real tasks from our projects. Here's where each one wins.
AI Agents Are Not Ready for Production (With One Exception)
The hype is ahead of the reliability. But there's one pattern where agents actually work today.
Your MVP Doesn't Need AI (Yet)
Ship the core loop first. Validate demand. Then sprinkle intelligence.
Hybrid Search Beats Pure Vector Search. Here's the Proof.
BM25 + semantic + reranking. We benchmarked it on 4 client datasets.
Running YOLO on Edge Devices: Lessons from the Factory Floor
Jetson Nano, INT8 quantization, and the latency budget you actually need.
Prompt Engineering Is Software Engineering Now
Version control your prompts. Test them. Review them in PRs.
Building a Document AI Pipeline That Actually Scales
OCR, layout detection, structured extraction, EHR integration. 14M+ docs/year.
The Build-vs-Buy Framework for AI Features
When to use an API, when to fine-tune, and when to build from scratch.
Stay sharp on production AI.
No fluff. Practitioner perspectives, once a month.