Jev + Treg is a crazy combo Already made 3 workflow saved us $8k/m: 1/ Fraud detection & website signup screening 2/ Triage buying signal 3/ Monitor viral content Literally drove business automation cost down to $0 Copy my automation flow at Fully open-source, repo link below 👇

Github:

So useful workflows and also how they're presented is super clean.

Thanks man!’

That's awesome Jason, love it. Big fan of Jev. Just discovering Treg. GTM simplified🔥 Creating you a spotlight on today.

Awesome thanks Said! Keen!!

@jasonzhou1993 the fraud detection bit is a lifesaver. Hit similar issues myself and custom screening totally changed the game for me too.

hadn't even heard of treg before glad i caught your post with the guide

thanks! Keen to hear your feedback!!

clean one!

GTM automation cost is driven down to $0 With tools like Jev & @treg_ai

for me Jev + treg very underrated combo will try it asap

Keen to hear your thoughts!!

yeah brother, soon

$8k/mo sounds impressive, but curious what deal/lead volume that's based on — could be overkill infra for a small team, or exactly right for a real pipeline. Triage buying signal seems like the most useful of the three; the other two feel doable with simpler tooling.

The cost story is compelling, but the durable metric is false-positive review load: a bash workflow that floods sales with bad signals is not automation. LangGraph (18k+ stars) is useful for making checkpoints and human handoffs explicit.

Jev does work with lab graph. But also I don’t see why can’t you just hand role the script instead of using man graph

solid work. confidence‑based routing keeps spam low while the cost hits zero

8k saved a month is a lot if legit

3 workflows in and the interns are getting nervous

😂 even managers are shaking!

the interns were just the warm-up

go Treg!

YOO, its EPIC 🔥🔥

saving $8k a month is impressive, but I'd like to know how sustainable that really is

For context we were losing $8k one month for credit farming, and we blocked bot & fraud using Treg + Jev

So Treg finds the data / evidence you need for Jev to make a judgement and assign a probabilistic value that your tools can use in their “routing” logic? … am I getting it ?

yep exactly, give Jev rich context and it decide & classify

The interesting part is the separation between detection and execution. Jev can handle the reasoning while Treg becomes the execution layer, which makes the $0 claim much more meaningful if the workflows actually run without human intervention. How much of the $8k/mo savings is from replacing manual work vs replacing paid tools?

im so excited to try these gtm recipes

fraud screening is where calibration has an actual dollar value. are you routing low-confidence cases to review or letting the typed decision hard-stop signups?

Yea we have a threadfold for review

learn from the master, just gave the open source project a star!

每月省 8000 这数字也太猛了，维护成本算进去了吗？

This is prob the best Jev tutorial so far

🫡🫡🫡

driving automation to $0.00 is real when you ditch heavyweight orchestrators. for our media pipeline targeting brazil, hand-rolled python scripts + sqlite state locks + local neural tts beat complex graph frameworks. boring linear steps keep costs at literal zero

It literally saved us $8k/m from credit farming bot on the jev <> @treg_ai workflow

@treg_ai Finally

The ability to route based on confidence score is really big unlock

Exactly, also the speed & cost is crazy