ExpertiseClientsServicesCRAFT methodContact
Book a call →
Case study · Hanny · 2026

Hanny, the AI community manager for freelancers and small businesses.

Brutux Studio designed and built Hanny, an AI that reads your website, learns how you speak and prepares your LinkedIn, Instagram and Facebook posts every month.

Hanny: “My posts” screen with a post’s text, its hashtags and its generated visual (sample brand)
SaaS · Generative AI · Social media
0→1
product designed and built by the studio
3
networks: LinkedIn, Instagram, Facebook

Product strategy · UX/UI · AI-assisted development

The context.

Hanny is a Brutux Studio product, launched in 2026. It is built for freelancers, very small businesses and SMEs who know they should post regularly on social media but have neither the time nor the appetite to spend their Mondays on it. Hanny plays the part of a community manager: it reads the company’s website, learns how it speaks, builds the month’s editorial calendar and writes every post with its visual and hashtags.

The starting observation is simple: almost everyone keeps it up for a few weeks, then the rhythm breaks. You have to find a topic, write, make an image and pick hashtags, every week, on top of everything else. Generic writing tools do not help: they produce smooth texts nobody dares to sign, and people end up rewriting them.

Victor Michel, the studio’s founder, carried Hanny end to end: product strategy, UX/UI design and development, done with the help of AI. The product currently covers LinkedIn, Instagram and Facebook, carousels included. The person stays in charge: they review, correct, then publish.

The challenge: There are plenty of AIs that write posts. The challenge was to build one that writes like the person, not like everyone else: a real editorial line across the month rather than a pile of posts, visuals faithful to the brand’s identity, and a review step simple enough that a busy business owner actually does it. Without ever letting out content nobody has approved.

The approach.

  1. 01

    An onboarding that reads your site for you

    Hanny starts from the website address. It works out the business, the audience, the tone, the topics, the colours and the posting times, then asks one question at a time, with the answer already suggested. You confirm or correct. Nobody wants to fill in a long form before seeing any result.

  2. 02

    Planning the month before writing

    You set the days, the times, the networks and what matters this month: a workshop, a closure, a launch. Hanny chooses the angle of each post with the whole month in view, then writes with what came before in mind. The calendar can be corrected before a single line is written.

  3. 03

    Writing in the first person

    The user specifies who is speaking: name, role, background. Every text is written in the first person, with a lived detail rather than a generality. Hanny then reviews its own output and reworks whatever rings false, to avoid the polite, interchangeable tone of generated text.

  4. 04

    Visuals in the brand’s exact colours

    Visuals are composed by a layout engine, not improvised: the right logo, the right colours, the right font, taken from the website. Each network gets its own dimensions, and a topic too rich for one image becomes a carousel. Photographs come from the brand’s library or are produced for the topic.

  5. 05

    Keeping the human in charge

    Nothing goes out without approval. You fix a word, swap the photo, try another layout, share a link with a partner for their opinion. Along the way Hanny tracks how much of the text gets edited: if it writes badly, that shows up in facts, not impressions.

The screens.

What this project taught

Hanny confirms one of the studio’s convictions: the value of a generative AI product lies less in the model than in everything around it. Upfront framing, the structure of the month, visual composition rules and the review step make the difference between a text you publish and a text you rewrite. Designing for AI means first designing for the person who will sign the result.

What it means for your project

If you are considering a product built on generative AI, Brutux Studio knows how to move from an impressive demo to a tool people use every month: framing what the AI must produce, keeping the user in the loop, measuring the real quality of the results. This experience serves building a new SaaS product as much as integrating AI into an existing tool.

(Let’s work together)

Your SaaS deserves better than “it works”.

30 minutes to talk about your product or your AI project. I’ll tell you honestly whether and how I can help.

Book 30 minutes ↗