Figr AI: UX Agent for Product Teams — Learns your product. Thinks through UX
Learns your product. Thinks through UX
Promoted
Maker
📌
Hey Product Hunt! I’m Matthew, founder of Epsilab, a Vibe Trading platform.
I’ve spent months wiring data, building pipelines, and babysitting backtests just to refine a single profitable strategy. That frustration is why I built Epsilab: go from idea to strategy in seconds without losing rigor.
Type a prompt like: “Create a buy only momentum breakout strategy for airline stocks. Buy more as the airline stock deviates higher than the basket mean.”
Epsilab turns that into an executable strategy and a full tearsheet in seconds. Iterate in plain English and see updated results instantly.
We’re bringing hedge-fund-level infrastructure to everyone: reproducible research, look-ahead safeguards, calendar-aware data handling, clear data windows, and optional walk-forward testing. Private by default; share when you’re ready.
What’s next: team workspaces for shared research and the ability to copy the best quants into your own lab so you can study and adapt their approaches.
We’re offering a free trial so you can start trading today!
Report
Congrats on the launch, Matthew! This looks incredible.
The ability to go from a plain English prompt to a full, executable strategy in seconds is an absolute game changer. I know people who spend weeks, if not months, on that exact process. You've really nailed the core frustration of quantitative trading.
Question: What was the most surprising or complex type of prompt that Epsilab was able to successfully translate into a viable strategy during testing?
Amazing work, can't wait to see this take off!
Report
Maker
Thank you @maxwowo! Generally, strategies that expect many hardcoded values or specific patterns that are incompatible with the interface we designed result in worse outcomes. Prompts that outline the logic, steps, or process with some flexibility generally lead to higher quality strategies (plus they're better when it comes to hyperoptimization).
Hey Product Hunt! I’m Matthew, founder of Epsilab, a Vibe Trading platform.
I’ve spent months wiring data, building pipelines, and babysitting backtests just to refine a single profitable strategy. That frustration is why I built Epsilab: go from idea to strategy in seconds without losing rigor.
Type a prompt like:
“Create a buy only momentum breakout strategy for airline stocks. Buy more as the airline stock deviates higher than the basket mean.”
Epsilab turns that into an executable strategy and a full tearsheet in seconds. Iterate in plain English and see updated results instantly.
We’re bringing hedge-fund-level infrastructure to everyone: reproducible research, look-ahead safeguards, calendar-aware data handling, clear data windows, and optional walk-forward testing. Private by default; share when you’re ready.
What’s next: team workspaces for shared research and the ability to copy the best quants into your own lab so you can study and adapt their approaches.
We’re offering a free trial so you can start trading today!
Congrats on the launch, Matthew! This looks incredible.
The ability to go from a plain English prompt to a full, executable strategy in seconds is an absolute game changer. I know people who spend weeks, if not months, on that exact process. You've really nailed the core frustration of quantitative trading.
Question: What was the most surprising or complex type of prompt that Epsilab was able to successfully translate into a viable strategy during testing?
Amazing work, can't wait to see this take off!
Thank you @maxwowo! Generally, strategies that expect many hardcoded values or specific patterns that are incompatible with the interface we designed result in worse outcomes. Prompts that outline the logic, steps, or process with some flexibility generally lead to higher quality strategies (plus they're better when it comes to hyperoptimization).
Chord - AI Group Chats
Love it - I was waiting for something like this, I've just been hacking together my own tools to do something similar.
@alexkayaian Thank you! Would love any feedback!