How We Fixed a Prompt That Gave Every AI Report an EV Section
Batch LLM report generation with zero failures, yet a prompt tuned on one industry gave all 201 reports an EV section. How we caught it and what fixed it.
Utkarsh Gupta
llm
Utkarsh GuptaHow We Fixed an AI Eval That Said 50% When the Model Was 97% Accurate
Our eval accuracy drop to 50% wasn't the model. Half the golden set pointed at retired IDs. How we remapped 1,009 items by content hash and fixed the labels.
Utkarsh Gupta
How We Tag New Documents Within Minutes Using an LLM and a Fail-Open Validator
How we built an LLM classification validator: one model proposes labels with reasons, a second votes keep or drop, and a broken vote keeps every label.
Utkarsh Gupta
How We Cut a Week-Long PDF Backfill to Three Days by Skipping Page Rendering
We stopped rendering PDF pages to images and sent PDFs directly to a vision LLM in 40-page parts. A week-long backfill finished in three days. Here's the cost.
Utkarsh Gupta
How We Turn 500-Page Documents Into Short, Grounded AI Articles
LLM article generation from long documents: how we turn a 591-page file into grounded articles, drop repeat stories by meaning and keep IDs stable across runs.
Utkarsh Gupta
How We Built an LLM Document Classifier That Labels Every Document in One Call
How we built an LLM document classifier: one model call returns five fields, label rules written as decision tests, our own schema check and a weekly eval.
Utkarsh Gupta