I built and shipped an AI lead qualification system in 1 week — here's exactly how it works
Here's the honest story of how it happened.
The origin
I built this originally as a client project — a solar engineering company needed a way to handle inbound leads without manual follow-up. I built the whole system in n8n, it worked, and I realized it was generic enough that any small business with a lead problem could use it.
So I packaged it up and put it on Gumroad for $39.
What the system actually does 6 workflows running together:
1. A lead comes in from a form, Gmail, or WhatsApp
2. GPT-4o-mini scores them 1-10 (Hot/Warm/Cold) based on budget, urgency, and need
3. Hot leads get a personalized email + Calendly booking link within seconds
4. If no reply after 48 hours, a personalized follow-up fires automatically
5. Everything syncs to HubSpot CRM and Google Sheets
6. Every Monday, an AI-generated pipeline summary posts to Discord
Zero manual input after setup. Runs 24/7. Costs under $5/month in API calls.
What I learned building it
The hardest part wasn't the AI — it was the edge cases. Twilio sends WhatsApp sender numbers as whatsapp:+91XXXXXXXXXX and if you don't strip that prefix before passing it to HubSpot, everything breaks silently. Took me an embarrassing amount of time to figure that out.
The duplicate detection is also basic right now — just an email match against Google Sheets. A proper implementation would use HubSpot's native dedup. That's a v2 problem.
Honest numbers so far
Revenue: $0 (launched today)
Gumroad listing: live
LinkedIn post: live
I have no audience. This is my first time selling anything online. Posting here because IndieHackers is the right community to share this kind of build-in-public journey.
Links
Full architecture on GitHub: https://github.com/AviVAvi/AI-Lead-Qualification-Sales-Automation
Workflow pack on Gumroad: https://asmitbohra.gumroad.com/l/eoyaey
Would genuinely love feedback on the product, the pricing, or anything else.
This mirrors my own experience closely.
I'm a doctor who built a free AI travel app
as a side project — same situation:
launched today with $0 revenue and no audience.
The "built it for a real use case first"
approach clearly works. My app came from
genuinely needing a better travel planning
tool while traveling.
One question: did you consider a freemium
model before going paid on Gumroad?
I went free + affiliate revenue, but I'm
second-guessing whether that was the
right call at this stage.
Congrats on the launch — this is actually a really solid use case of GPT + automation.
I work on Egyptian Arabic conversational datasets that help improve real-world AI interactions (especially multi-turn dialogue and decision flows).
If you ever expand to multilingual users or want to improve conversational quality, I can share a small sample for testing.
Shipping this in a week is solid — especially with the edge cases you mentioned, that’s usually where most people get stuck.
What stood out more is you said this is your first time selling.
Out of curiosity — are you seeing more traction from the “automation system” angle, or from the specific outcome (like faster response / higher conversion on inbound leads)?
Feels like that framing alone could make a big difference in how people perceive it.