@brycent_ — 9x breakout selling ReasonBlocks

Loading video…

This Startup Makes AI Cheaper and More Reliable ✅

@brycent_

151.2K followers · usually 5.1K views a post

Views

43.6K

Likes

2.0K

Comments

162

Promoted product

ReasonBlocksreasonblocks.com

Agents that remember how they solved problems. Less time on bugs, less money on tokens.

Why it worked

The video effectively showcases a clear value proposition (cheaper and more reliable AI) with quantifiable results (52% token reduction, 42% accuracy lift). The founder's personal story of starting young and getting into Y Combinator adds credibility and relatability, while the explanation of the technology is concise and easy to understand.

Summary

The founder of ReasonBlocks explains how their AI runtime layer makes AI agents cheaper and more reliable by reducing token usage and improving accuracy. He details the technology behind it, including their trained model for identifying important information and a system for reusing past successful agent runs. The founder also shares his background, starting with research and publishing papers before dropping out of Stanford to pursue the startup, which he then got into Y Combinator.

Free account

Transcript, structure and on-screen text

5 beats, a 493-word transcript and 78 lines of on-screen text — the parts you need to write your own version.

Original caption

This Y Combinator-backed startup is building a runtime layer that makes AI agents cheaper and more reliable when using them. Startup - Reason Blocks (reasonblocks.com) Reason Blocks is a runtime layer that makes AI agents cheaper and more reliable. They correct agent reasoning and compress tokens mid-run, which results in compounding gains across runs. On SWE Bench Pro, ReasonBlocks has seen a 52% token reduction and 42% accuracy lift. Startup founders pitch Brycent at Y Combinator Demo Day.