ScanMessage — AI fraud detection for WhatsApp & SMS in Latin America
Hey IH! Just launched ScanMessage (https://www.scanmessage.dev/) after a few weeks building it in my spare time.
The problem: fraud via WhatsApp and SMS is massive in Latin America. Existing tools detect it but don't explain it — and most are built for English-speaking markets.
What ScanMessage does: you paste any suspicious message and the AI returns a risk score, detected signals, and a plain-language explanation of why it's dangerous and what to do.
Biggest learning so far: fraud scoring requires signal co-occurrence. A single signal like 'requests bank info' isn't enough — you need urgency + impersonation + suspicious link + time pressure together before assigning a high risk score. Took a while to calibrate.
Currently free, B2C first to build real fraud data, then B2B targeting fintechs.
Would love feedback from anyone who's tackled fraud detection or B2B sales in emerging markets.
Strong problem, especially if you’re moving toward fintechs later.
One thing I’d watch early: ScanMessage explains the function, but it still feels like a small consumer utility.
The actual product is bigger than that.
You’re not just scanning messages.
You’re building a fraud signal layer for WhatsApp and SMS in Latin America.
That matters because fintech buyers won’t only judge detection quality.
They’ll judge whether the product feels trustworthy enough to sit near risk, compliance, and customer protection.
ScanMessage works for B2C testing.
But if the long-term play is B2B fraud infrastructure, the name may start feeling too narrow and lightweight.
Davoq.com would fit that direction much better if you ever want this to feel like serious risk/security infrastructure instead of a message scanner.
Thanks for the thoughtful feedback — really appreciate it.
You're right that the real play is bigger than a message scanner. The long-term vision is exactly what you described: a fraud signal layer for WhatsApp and SMS in Latin America, with B2B as the main direction.
Right now I'm in the first stage — B2C to build real data and validate the detection quality before approaching fintechs. The name works well for that phase.
The positioning shift toward infrastructure is something I'll keep in mind as the B2B conversations start. Good flag on how fintech buyers think about trust and compliance.
Will definitely revisit as the product evolves.
That makes sense for validation.
But I’d be careful with waiting too long.
If fintech is the real direction, the trust frame starts before the first B2B conversation.
Because when you show early traction under “ScanMessage,” buyers will still mentally file it as a consumer scanner first.
That creates extra work later:
prove the tech
then reframe the category
then rebuild trust around infrastructure
If the product is already collecting fraud signal data, the infrastructure story is already starting now.
B2C can be the data engine.
But the brand should probably start preparing for the B2B trust layer earlier than the sales motion does.
That’s where a name like Davoq makes more sense — not as a cosmetic rename, but as a cleaner risk/security frame before fintech buyers anchor you as a lightweight scanner.