Private beta · for marketers, designers & researchers

Your users don't lie.
They just can't be honest.

Social pressure, cultural barriers, and personality bias distort your research, especially in the concept definition phase, when honest feedback matters most. AI personas have none of those filters.

A user research interview: the participant says one thing but thinks another

The problem

The problem isn't dishonesty.
It's human nature.

In focus groups and user interviews, participants want to be helpful. They don't want to seem negative or hurt the researcher's feelings. Cultural norms, social anxiety, and personality type all push responses toward the positive, regardless of what they actually think.

This bias is most damaging in the concept definition phase, when you most need signal and least need noise. AI personas don't have social anxiety. They don't have cultural filters. They just react.

Focus group: participants give thumbs up while thinking brutal thoughts

How it works

Pick a persona. Start talking.

Voice

Real-time voice conversation

Talk to your persona as you would in a real interview. They interrupt, push back, and ask the questions your real users never would.

Chat

Text-based feedback

Prefer typing? Chat mode gives you the same brutal honesty in writing. Share a link, a screenshot, a concept, and get a reaction.

Mobile

Works on any device

No app to install. Open it on your phone during a commute, on a tablet in a workshop, or on your laptop between calls.

Alex

Alex

Canadian student · Milan

CHAT
We're building an app that matches foreign students with verified landlords in Milan. What do you think?
I mean... I already use Facebook Marketplace and it works. Why would I switch? What's actually different here?
We verify landlords and handle contracts digitally.
Okay but who does the verification? And what happens if something goes wrong after I move in? That's the part nobody ever answers.

The conversation

No filter.
No social pressure.
Just feedback.

Each persona has a backstory, a set of frustrations, and zero incentive to spare your feelings. They challenge your assumptions, poke holes in your logic, and ask the questions your real users never voice.

You can set the Brutality Level, from Mom (your idea is brilliant) to Fatality (the internet troll treatment), before each session.

The interlocutors

11 personas. Zero filters.

Alex

Alex

Canadian student · Milan

Andrea

Andrea

Brazilian student · Polimi-bound

Jan

Jan

Swedish consultant · frequent traveler

Marco

Marco

CEO · Nesthive Italia

Prof. Ferretti

Prof. Ferretti

Rector · Università di Milano

Roberto

Roberto

Operations · CasaStudenti

Elena

Elena

CEO · Valorian Hotels Italia

Gianluca

Gianluca

Airbnb Host · Milan

Liis

Liis

Estonian traveller · solo

Kavya

Kavya

Indian student · Milan

Rinaldi & Venturi 🇮🇹

Rinaldi & Venturi 🇮🇹

CMO & Head of Brand · Game Training

We don’t believe synthetic users replace real ones. But in the early stages of concept definition, the social pressure, personality, and cultural barriers of real users often make them less useful than a well-designed AI persona. Not because they’re lying. Because they’re human.

The Brutality Level

You choose how much truth you can handle.

Drag the lever. Same persona, completely different energy.

👌Level 2: Honest
We're building an app to match students with verified landlords in Milan. What do you think?
I already use Airbnb and it works fine. Why would I switch? What does 'verified' actually mean here, and who's doing the verification?

For researchers

Create personas tailored to your research

Beyond the 10 public personas, researchers can create custom personas specific to their study context. Write a system prompt, and the platform automatically generates a portrait and assigns a voice.

  • Write a system prompt with identity, personality, and context
  • Avatar generated automatically in the platform's illustration style
  • Personas are private, only visible to your group
  • Unlimited personas on the Research plan
A designer reviewing a prototype: saying one thing, thinking another

Two ways to use Brutal Feedback

Same engine. Different goals.

Whether you're validating a new product idea or training your team to pitch better, Brutal Feedback has a mode for you.

For Product Development

Validate concepts before you build.

Talk to AI personas that represent your target users — students, business owners, frequent travelers, hotel managers — and get the unfiltered reactions you'd never hear in a real interview. Ideal for product designers, UX researchers, marketers, and startup founders in the early concept phase.

  • 10 public personas across student, business, and traveler segments
  • Voice and chat modes for flexible research sessions
  • Brutality Level slider from diplomatic to fatality
  • Custom personas for your specific research context

Best for

Product designers · UX researchers · Startup founders · Marketers

For Presentation Training
Beta · Italian only

Train your team to pitch like executives.

A dedicated rehearsal mode for pitch and presentation training. Perfect for companies, business schools, and coaches who want to give their teams a safe space to practice executive communication — and get scored on it.

The scoring system

Clarity

Is your message clear and structured?

Confidence

Do you own the room?

Relevance

Are you speaking to the right priorities?

Handling objections

Can you hold your ground under pressure?

  • Rinaldi & Venturi: a CMO and Head of Brand who score your pitch
  • Ideal for companies, business schools, and executive coaches
  • Designed for team rehearsal — not just solo practice
  • More languages and personas coming soon

Best for

Sales teams · Business schools · Executive coaches · Startup accelerators

Private beta · registration requires admin approval

Ready to hear the truth?

Join the beta and test your concept with personas that have no reason to lie to you.