Activity
Mon
Wed
Fri
Sun
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
What is this?
Less
More
9 contributions to AI Automation Agency Hub
Has anyone used REA (Reverse Engineer Anything) yet? What have you used it for?
Hey everyone, I'm about to start using REA, the skill that's blowing up on GitHub. For anyone who hasn't seen it: you install it with one command in Claude Code, Codex or Cursor, and it lets the agent analyze how an app, a website, a desktop app or an APK works, and it comes back with what it found plus the evidence, instead of guessing. My idea is to use it to better understand tools I already use in my business and in client projects: how a site talks to its backend, why an integration is failing, or how an app is put together before I connect an automation to it. I'd love to hear from people who are already using it and from those thinking about it: What use cases are you using it for? Which agent does it run best with, Claude Code, Codex or something else? Any precautions you take with permissions or security, since it runs locally with your access? Any case where it didn't help or fell short?
Where would you start with lead gen systems if you've never built one?
Hey everyone, I'd love some guidance on this one. I keep seeing that lead generation is one of the easiest automations to sell, because the value is clear to any business owner: more qualified conversations. I want to start building lead gen systems for clients, but I'll be honest, I have zero experience with it so far. So I'm curious how you'd approach it if you were starting from scratch: What's the first lead gen system you'd build to learn and to show as a portfolio piece? Is it better to start with outbound (scraping, enrichment, cold email) or inbound (capturing and qualifying leads from forms, ads, or WhatsApp)? What stack or tools would you learn first, and which would you skip for now? Any videos, repos, or templates that helped you get your first one working? For those of you already selling lead gen, what does a simple first offer look like, and which niches have been the easiest to start with? Thanks in advance, I'll share what I end up building so we can learn from it together.
3 likes • 26d
@Rob Olson Thanks Rob, that's a lot of real experience in one comment. Splitting it by ICP, outbound for B2B and inbound for B2C, is a really useful way to look at it, and learning with my own money before touching a client's budget makes a lot of sense. My niche is tourism in Costa Rica and Latin America, so the people booking are mostly travelers, and a simple landing page with a short quiz into a sheet is something I can test early. Saving the tip about studying simple niche funnels too.
3 likes • 26d
@Harsha Agarwal Thanks Harsha, agreed. Form to CRM to follow-up is where I'm starting, and being able to measure it from day one is a big plus. Outbound can wait until I have that flow down.
Where would you start if a client asked you to build a CRM?
I have a client who wants a CRM, and I’m scoping the ground before I start building. The only clear first move I have so far is interviewing someone who lives in the current process day to day: what they do today in the CRM (or the spreadsheets standing in for one), what hurts, and what they cannot lose. From that I want to ship something that solves their case and also becomes a reusable template I can offer other clients without starting from zero every time. If you’ve built a CRM or heavily customized one, where would you start? What would you learn before writing code? Any videos, repos, stacks, or templates that actually shortened the build for you? Concrete tips welcome. Thanks.
0 likes • 29d
@Malik Ahmed Source of truth is the part I'd underline too. For this one it ended up being a single service list for rates, instead of prices typed into every deal and template. And before anyone switches, I'm checking the migrated data deal by deal against the old system's export.
0 likes • 29d
@Lakshya Pandey They're on Pipedrive today, not spreadsheets, and they'd already been moving follow-ups out of it into other tools. What they need turned out to be narrower than a full CRM: email synced per user and grouped by deal, plus a few pipelines. On reuse I'm with you, it's one product and the client-specific stuff goes in config, not in the code.
Is the recording from few days ago webinar available anywhere?
I'd love to rewatch the part where they walked through the step-by-step process for auditing a business and spotting automation opportunities. That framework was the most useful thing I've seen here in a while and I want to study it properly instead of relying on my notes. If there's a replay link (or even just the slides), I'd really appreciate a pointer. Thanks!
1 like • Aug 20
https://live.aaaaccelerator.com/recording Here is the link for everyone that need it like me, there is 3 days left before its gone so record it and save it because there so much value there and hopefully we can put it into action
0 likes • Sep 25
@Malik Ahmed Appreciate you saying that, Malik. Let me check where the replay and slides landed and I’ll share what I can find here. The business-audit / automation-opportunity framework is worth revisiting, so I’ll try to get you something useful.
How are you planning to use Jev in your workflow?
Curious how people here are using Jev, or thinking about using it. If you already have it in your stack, or you are planning to add it, drop a use case below. What part of your workflow does it cover, and what changes after you add it? One idea I want to try: use Jev as a cheap decision layer in front of my agent stack. Incoming client messages and ops tasks get classified first (which client, how urgent, whether it needs a human before anything goes out). That way the heavy models only run when the decision is worth it, and risky sends stay gated. I want this thread full of real examples we can learn from and adapt to our own setups.
1 like • Sep 21
@Bharti Kaith Thanks for sharing this, Bharti. The hotel and lead qualification examples map well to a pre-classifier. Simple FAQs can stay lightweight, while booking actions and ambiguous messages can route to the heavier agent or a human. Which flow would you test first?
0 likes • Sep 25
@Malik Ahmed That’s exactly the lane I’m using it for, Malik. Cheap decisions first (classify, route, urgency, human-before-send), then only call the heavier agents when it’s worth it. Makes the escalations more predictable and keeps the expensive models off the easy cases. If you’ve already wired something similar, I’d love to hear what you put in the typed payload.
1-9 of 9
Kendal Segura
4
52 points to level up
@kendal-segura-9302
Learning every day

Active 19h ago
Joined Jun 13, 2026
Powered by