Building a backtesting engine that never stops searching. 24/7.
@RHerman14K followers
18,302 views
38 Comments
Comments from the original post on X (Twitter)

@Bentrades_ttfm
🥹🔥
3 likes

@RHerman
The next generation of backtesting 💪
1 like

@Bentrades_ttfm
You’ll have my fully support about this!! 💪
1 like

@RHerman
Thanks Ben! Means a lot. Expect more in the future!
1 like

@Bentrades_ttfm
🫡
1 like

@str0ngl4nd
I'm actually building this too. Different stack and my approach is different than yours but yes... working on it.
2 likes

@RHerman
Took me a couple of days - and still improving it. Let me know how it goes

@MacFlyMM
how do you make it search 24/7
2 likes

@vfnoca
Spent all last year building mine. Would suggest trying other cheaper models or the costs will be a bottleneck. Had success with Google’s models
2 likes

@RHerman
Ohh dont mention that :) Cost of the business.

@spikepanel
👏
1 like

@burritolnx
Great work man, congratulations
1 like

@RHerman
Thanks buddy. Still working on the engine - its a beast already.
1 like

@peanutchunky21
goat
1 like

@Rebound_Trader
We’ve been building a very similar loop inside TheSystem Next — but the interesting part for us starts after the idea is found. Research → strategy generation → backtest → optimization → walk-forward / out-of-sample validation → robustness filters → reject or promote. The goal isn’t to find as many “profitable backtests” as possible. It’s to automatically kill fragile ideas fast, keep statistically defensible ones, and continuously repeat the process across markets. 24/7 discovery is powerful. 24/7 falsification is probably even more important. @thesystemnext
1 like

@Crypto_anarchie
Beautiful setup! What platform do you do the backtests on?
1 like

@RHerman
Thanks. OHLCV file from databento

@onstocks_india
can yu share more detials
1 like

@RHerman
SUre- will post a more videos
1 like

@onstocks_india
Any further updates
1 like








