Why we built it
We've built support tools for a long time. The pattern is always the same: ten percent of questions account for ninety percent of tickets. Documentation answers most of them - but only if visitors find the right page, and read it, and apply it to whatever screen they're stuck on.
Most AI chatbots can read documentation. They can summarize, paraphrase, and link. But when a visitor asks "where do I export my invoices?", a chatbot answer like "Click the Export button in the top right of the invoices page" is one navigation and one extra cognitive jump away from being useful.
We thought: what if the assistant could just point at the button? What if, with consent, it could even click it?
That's Bruno.
What we believe
- Helpful beats clever. A short, concrete answer with a highlight beats a paragraph of paraphrase.
- Consent is non-negotiable. Bruno asks before it acts, every time, with a preview. If we wouldn't be comfortable with this happening on our own bank's website, we don't ship it.
- Honest about uncertainty. If Bruno doesn't know, it says so and offers a sensible fallback - a docs link, an email, a human.
- The page is part of the prompt. The visitor's current page is the most useful context we have. We use it carefully and we never read what we shouldn't.
- Tenant data stays in tenants. No knowledge bleed across customers. No training on your content. No clever exceptions.
How we work
Bruno is a small team. We ship continuously, write things down, and answer help@bruno.support ourselves. If you find a bug, the person who fixes it is probably going to be the person who replied to you.
Get in touch
We're always happy to hear from teams using Bruno or thinking about it. Email help@bruno.support and you'll get a real reply.