Pinned
Ramp Labs
90 posts
We got early access to @thinkymachines’ Tinker and have been experimenting with it this week at Ramp Labs. We used it to compare RL post training performance between an ensemble of domain specific models vs. a singular model. It handled infra, model updating, and GPUs smoothly- Meet Agent Fill, our agentic form filler. Built for finance & ops teams that don't want to waste time filling out PDFs. Available today in alpha.
00:00 - The sharpest teams don't fill out PDFs. They deploy agents. Coming soon from Ramp Labs.
GIF - Introducing LLM Games 🕹️ by Ramp Labs We pit GPT-5, Grok-4, o3, Gemini-2.5, and other models against each other to play Connect Four. GPT-5 high crushed all models – winning 14/14 games. As the games progress, the models think a lot longer. View the full replays of the games
GIF - Agent Fill (our agentic PDF filler) just got smarter: - Hover to see exactly why each field was filled - Review and complete missing fields directly in the editor We've been working with finance and ops teams to save hours every week - thank you all for the feedback. Try it
00:00 - Replying to @RampLabsWe're also hiring AI Product Engineers at @tryramp. Join us to build the future of finance AI. jobs.ashbyhq.com/ramp/d4c1b7f2-…
