Why Reducto Built Horizontal Primitives for Document AI
Reducto co-founder Adit explains why solving messy medical faxes creates stronger document AI models than building vertical parsers.
Reducto Final YouTube 1080p Sandwich Title
With Adit · Friday, November 6, 2026
In this cooking-show format interview, Reducto co-founder and CEO Adit discusses the evolution of document parsing from early PDF drivers to agentic vision systems. He breaks down why general frontier models silently fail on recall, why Reducto transitioned to specialized in-house models like R1, and how the company evaluates document reliability in high-stakes industries like healthcare and finance. Adit also shares insights on expanding into document editing and localization, as well as Reducto's culture of hiring craft-obsessed, LLM-native engineers.
Key takeaways
- Vertical document extractors are easier to sell early, but they trap engineering teams into maintaining narrow, brittle rule sets. Read more →
- General frontier models fail silently on recall when parsing complex documents in high-stakes fields like finance and healthcare. Read more →
- Reducto CEO Adit argues white-collar work has not reached its agent milestone because tools stop at data extraction rather than editing live files. Read more →
- Frontier models with uncapped reasoning time beat specialized platforms on precision, but they fail silently on recall by dropping table rows and context. Read more →
4 articles from this episode
Why Reducto Abandoned Frontier Model APIs
Reducto co-founder Adit explains why task-specific models beat commercial LLM APIs on accuracy, latency, and cost at scale.
Reducto CEO Adit: Document Agents Must Write Back into Live Files
Reducto CEO Adit explains why white-collar AI agents must master visual layout manipulation and write back into documents to automate office work.
Why Frontier Models Drop Data in Document Extraction
Reducto CEO Adit explains why frontier models silently drop rows in document parsing and how vision systems fix recall.
More Latent Space episodes
- Mini Brains, Beating Heart Organoids, and the AI That Predicts Them — Fabian Theis, Helmholtz MunichNov 8 · 4 articles
- ⚡️ The 2026 Software Factory Standard — Zach Lloyd, Warp.devNov 8 · 6 articles
- Deep Dive: Anthropic S-1 Myths and Mythos — Jamin Ball, AltimeterNov 8 · 4 articles
- Behind Latent Space, ft AIEi Shanghai preview — with swyx and Galen GaoNov 8 · 5 articles