Mindlet
AI Customer Support That Works 24/7
Hey everyone 👋
I’m Umair, a solo founder currently building Mindlet.
Over the last few months I kept noticing the same thing when talking to small businesses and founders:
They spend a lot of time answering the same customer questions.
Things like:
• pricing
• booking availability
• product details
• delivery times
Most of these questions are repetitive, but they still require someone to reply manually.
When I looked at existing solutions, the problem became obvious.
A lot of customer support tools with AI features are built for bigger companies — and they’re priced that way too.
For many small businesses, paying $100–$300/month for support automation just isn’t realistic.
So they either:
keep answering everything manually
ignore messages after hours
or lose potential customers
That’s what pushed me to start building Mindlet.
The idea is simple:
An AI assistant that learns from your website or documents and answers customer questions automatically.
But the real goal is making it affordable enough for smaller businesses and solo founders, not just big companies.
Right now I’m focusing on:
• improving how the AI learns from website content
• making setup extremely fast
• keeping pricing accessible for smaller teams
I’m also starting outreach to get early users and real feedback.
I’ll share the progress here as I go — both what works and what doesn’t.
If you’ve built in the customer support / AI space before, I’d love to hear your thoughts.
— Umair
1 Comment
Nice direction! Repetitive support questions are a real pain for small teams.
Since Mindlet learns from website content and answers customer questions automatically, a few security points will matter early:
• If the assistant reads website content or uploaded documents, are those sources stored or only indexed temporarily?
• Are customer conversations or email captures retained, and for how long?
• If responses are generated through external LLM providers, are prompts and user messages retained or used for training?
Public chat assistants also face abuse quickly:
• prompt injection attempts
• automated spam messages
• scraping of internal knowledge sources
Rate limits and strict boundaries on what the assistant can access will help prevent that.
Another trust signal is transparency. Small businesses will want simple answers:
• what data is stored
• how long conversations are kept
• whether AI providers train on their content
As a security team building Nautillo Pro, we often test public chat widgets and support bots because they sit directly on exposed website surfaces.
We also offer a free version.
If you want to evaluate how your platform behaves from an external attacker perspective, Nautillo Pro is available to test.