Lukas Kelsey-Friedemann
I'm the founder of Well, an analysis and automation tool for immigration lawyers.
I previously went through the YC W23 batch. First building Cherrry, a semantic search developer tool. Then, Readlet a text to speech app for learning quickly.
Before that I was a software engineer at Roboflow (YC S20).
Off the computer, you'll find me on a bike ride or on a walk with my wife and dogs.
You can find my resume here
Work
Y Combinator
Winter 2023 Batch
After being accepted, I built Cherrry, a semantic search developer tool. Making vector search easy for every developer without ML/AI knowledge. Later during the batch I met with my group partner, and we decided it would be better to pivot the company. I explored lots of ideas, business models, and potential markets.
Well
Founder
Well is an analysis and automation tool for immigration lawyers. We extract data from client documents and automate client intake, form filling, and document generation.
Readlet
Founder
Readlet is a text to speech app for learning quickly. You can upload any text and it will be converted to audio with life like voices. Ultimately, I decided to pivot the company away from this idea after failing to gain enough early traction.
Cherrry
Founder
Cherrry was a semantic search developer tool. Making vector search easy for every developer without ML/AI knowledge. A fully managed service that does all the text chunking, embedding, and search. Allowing you to use our open source SDK's and rest APIs.
Roboflow (YC S20)
Full-Stack Engineer
Lead the development and growth of Roboflow Universe, a platform for AI developers to share computer vision datasets and models. I also worked on Roboflow's core product, a suite of tools for training and deploying custom computer vision models.
Writing
Mixing Models: Hacking Improved Performance from AI
In Progress