Data for the Real World: founder validation breakdown
Original RFS by Austin Tindle and Diana Hu. Source: Y Combinator Requests for Startups.
The RFS in plain English
YC wants startups collecting and structuring dense real-world data so AI can understand physical environments as well as it understands text and code.
Why now
Foundation models are advancing, sensors are getting cheaper, and many physical-world systems still lack machine-readable ground truth.
IdeaCheck angle
This can be valuable if the startup owns a rare data source tied to a specific workflow, not just a broad data-collection pitch.
Founder trap to avoid
Data businesses often underestimate collection cost, labeling difficulty, rights, privacy, and how long it takes before the dataset becomes defensible.
A concrete idea to test
A sensor and data platform for construction sites that turns physical progress, safety risks, and material movement into structured AI-ready data.
Test your version
Value your startup idea
Working on something near physical-world data? Get a premium IdeaCheck verdict before you build.
Unlock the full verdict with premium IdeaCheck.