Floodlight Labs
An independent software lab that builds and runs its own products.
A San Francisco software lab working across web infrastructure, domain intelligence, verification, analytics, market research, and practical AI.
How it works
The portfolio is the company.
Every project listed here is designed, built, deployed, and maintained by Floodlight Labs. Some are public utilities, some are research systems, and some are paid products. All of them solve a specific problem for people working on the web.
The work includes monitoring newly registered domains, verifying files, filtering bot traffic, researching markets, and making small AI models useful in production. Products stay intentionally focused so they remain understandable, affordable to operate, and worth maintaining.
Current areas
- Emerging websites and domain intelligence
- File authenticity and verification
- Market data and research tools
- Analytics with aggressive bot filtering
- Local AI inference and automation
Built to be operated
Small systems, owned end to end.
Floodlight products run on infrastructure we understand and maintain: self-hosted Linux services, small cloud instances, compact data pipelines, and local language models on our own GPU hardware.
Keeping the stack close makes failures easier to diagnose, costs easier to control, and useful improvements faster to ship.
Field notes
What we learn gets written down.
The notes cover what the work actually reveals: which web signals survive contact with real data, where crawlers and bots distort the picture, how local AI behaves in production, and how far modest infrastructure can be pushed.
Read the notes →