I cold emailed 50 compliance officers about the EU AI Act deadline. Here's every reply I got.
In April I was reading about the EU AI Act and realized most fintechs using AI for loans and credit decisions have no audit trail. Not a real one. Their database is editable. A regulator can't trust it.
So I built AIDAL.
One API call after your AI model runs. Every decision gets a SHA-256 hash chained to the previous one — mathematically tamper-proof. Any regulator can verify it independently without logging in, without trusting us, using only math.
Here's the 3-minute demo: https://youtu.be/i9gs1EUyb80?si=hKREBJz2GFf8yTkM
What's live:
- SHA-256 hash chain with public verification page
- AI-generated plain English explanation per decision
- Compliance checks for EU AI Act, MAS FEAT (Singapore), OJK POJK 11 (Indonesia), VARA (UAE)
- One-click regulator-ready PDF download
- Model registry, fairness detection, incident reporting
- 30-minute integration
Honest numbers:
- 50+ cold emails sent, handful of warm conversations, 0 paying customers yet
- Built in 2 weeks solo
- Free during beta
What I've learned so far:
- "We're not that far ahead yet" is the most common reply from compliance officers. That's not a no — that's a not yet, and August 2 is the forcing function.
- Developers integrate in 30 minutes. The hard part is getting to the compliance officer above them.
- Building from Indonesia for Singapore/EU is harder on distribution, easier on cost.
58 days left. August 2 is the EU AI Act enforcement deadline. After that the urgency disappears and it becomes a slower sales cycle.
Free API key at tryaidal.com if you want to try it.
Happy to answer anything — architecture, regulatory details, what's working in outreach.
