Building products that solve real problems

Robin Bonduelle: Building Products that Actually Solve Problems

When people talk about AI meeting tools, it usually sounds the same. Notes. Summaries. Action items. Another bot in your calendar.

Robin Bonduelle took a different route. As Co founder and CEO of Claap, he is quietly building an AI product that started as “just” a better Loom and turned into something much more ambitious

a conversation operating system for revenue teams.

This is the story behind it.

From consultant to product obsessive

Robin did not start as a classic “born founder”.

Before Claap, he spent more than a decade as a product manager. Twelve years of shipping features, working with designers and engineers, and trying to answer the same question again and again

Does this thing we are building actually matter

Product satisfied something deeper for him. He was not a blue collar worker using his hands, but he still wanted to see a clear before and after.

Before there is no app.

After there is something people use every day.

There was also a more honest reason. He never fully clicked with managers. He liked to learn by doing, failing, and figuring things out himself. At some point, the only real way to test himself was to bet on his own product and his own ability to chase revenue, not just good design.

Claap became that test.

The founding team behind Claap

Claap has three co founders.

• Robin runs product
• One co founder leads marketing, after years at a fast growing edtech company in New York
• The third is a veteran from Seattle with fifteen years at Amazon, bringing deep execution and operational discipline

Together they mix consumer grade UX, strong storytelling, and the kind of operational thinking you only get from shipping at Amazon scale.

Claap started somewhere else

Claap did not begin as the product it is today.

The first version, launched in twenty twenty one, was an asynchronous meeting solution meant to replace parts of Loom. User growth was good, engagement was solid, but revenue lagged behind. Classic product first story.

While Robin and the team were trying to tighten the sales motion, they noticed a bigger pattern

they had a mountain of meeting recordings and no way to get real insight from them.

They were drowning in calls and doing almost nothing with the information inside.

That is the moment Claap really started.

Instead of being a nicer async video tool, it pivoted to something more interesting

a way to mine customer and internal conversations for actual signals, not just transcripts.

What Claap is now

If you look at the surface, Claap still feels familiar.

You can

• record calls through a bot
• record your screen through a desktop app or extension
• get transcripts in many languages
• receive meeting notes and summaries
• collaborate on comments and highlights

That is the “note taker” everyone knows.

The deeper layer is where it becomes different. Claap is moving from “what happened in this meeting” to “what is hidden across all our meetings”.

Think of it as Clay but for conversations instead of leads.

You can build smart tables on top of your calls and ask questions like

• Which prospects asked for a mobile app in the last twelve months
• What exact pain points did they mention
• Where do they push back on pricing

Claap parses your entire call history and fills structured columns with the answers. Those insights can go to your product team, your CRM, or into AI agents that act on top of the data.

It shifts AI from a personal assistant that writes a recap to a company memory layer that you can query.

Why Claap focused on sales rather than everyone

In the beginning, Robin tried to do what many first time founders do.

Horizontal product. Product led growth. Anyone can use it. Anyone can pay. Loom energy. Notion energy.

In reality, it created noise.

Designers loved it. Internal teams used it. But those personas rarely have budget or urgency. When the time came to monetize, a lot of “power users” vanished.

That is when Robin made a hard but important shift. Instead of trying to be the universal async tool, Claap focused the narrative and product on sales teams and go to market motions.

Sales teams

• feel the pain of messy calls and missing insights
• have budget
• directly connect to revenue

Internal use cases are still possible, but they are treated as expansion, not the core story. Message first, then everything else.

The French UX factor

A funny theme in the conversation

Dom pointed out that many French products feel unusually polished. Robin laughed but did not fully disagree.

In B2C, good design was always non negotiable. In B2B, for years, people treated UX as optional. That time is gone.

Companies like Typeform showed that interaction design can be a growth driver, not decoration. Claap follows the same belief.

But Robin is very clear on one thing. Design is not about fancy gradients and animations. It is about helping the product hit its north star metric.

Beautiful UX that does not help activation, retention, or revenue is just decoration.

Sometimes the right design is clean and minimal. Sometimes, like Amazon, it is not visually sexy at all but incredibly efficient for billions of people.

The point is not aesthetics for their own sake. It is whether the design improves the outcome that matters.

Integrating AI when your product depends on it

For Robin, AI is not a side feature. It sits in the core of Claap.

That makes building much harder.

Classical development is deterministic. For a given input, you know exactly what output the system will produce. You can write tests, run them, and be confident.

LLMs are probabilistic. Change one word in a prompt and you might get a different structure, tone, or missing detail. That is fine in a simple chat interface. It is not fine when the LLM output becomes input for real application logic.

Claap had to

• define strict JSON outputs
• build curated data sets of real meetings across languages and use cases
• run every prompt variation across those sets and compare results
• sometimes use one model to evaluate the output of another

They also learned they needed dedicated AI specialists. In the early days everyone wrote prompts and shipped small features. Later they realized they needed people who go deep into best practices instead of spreading AI knowledge thinly across the whole team.

Remote first and async first culture

Claap is fully remote and almost fully asynchronous.

From day one it made sense. They were building tools for async collaboration after all. It would be strange to be office only.

Some key things Robin highlighted.

Remote only works if you are intentional.

Nothing is implicit. You have to design everything.

• Hiring
• They filter for people who already succeeded in remote setups or clearly have the mindset for it. If there is doubt on either side, they do not hire.
• Working hours
• Freedom comes with expectations. Team members can take long breaks during the day, but they must communicate clearly and update calendars.
• Documentation
• Remote forces you to write. Specs. Decisions. Processes. It feels heavy at first, but it pays off in clarity and onboarding.

Claap also runs off sites twice per year. One week in a single house to brainstorm, run hackathons, and think beyond the local maximum. Robin is convinced you still need in person time to unlock creativity and long term direction.

On a normal week though, he has almost no internal meetings. One to two hours total. Everything else is handled through async updates and Claap itself.

Where remote breaks down

sales and juniors

There are two groups Robin is still figuring out.

Sales people and junior team members.

Sales teams often thrive on energy, physical presence, and immediate feedback. Several remote sales hires at Claap did not work out, even though they were experienced. Many came from big company environments where pipeline and structure already existed and struggled with the chaos of startup selling plus full remote.

Right now, Claap is moving to a hybrid “archipelago” model for sales, with hubs starting in Lisbon and possibly more locations later.

Juniors are the other challenge. Remote work is amazing for senior people with families and established careers. For someone twenty three who is trying to build skills, networks, and friendships, a lonely home office can be brutal.

Claap currently hires mostly experienced people. Robin is still not fully convinced how to give juniors the social and learning environment they need in a fully remote setup.

Hard founder lessons from a first time CEO

Looking back, Robin is very direct about what he would change.

• He would not start as a horizontal PLG tool trying to be for everyone
• He would not optimize for engagement without caring if those users would ever pay
• He would anchor on revenue from day one, even as a VC backed company

Claap raised three point five million on a deck. That made it easy to fall into the “grow usage, figure out monetization later” trap. It worked for Facebook. It works for a handful of consumer giants. It does not work for most B2B SaaS products.

Monetization is a filter.

The people who say they love your product and use it every day are not always the people who will pay. If you optimize only for their happiness, you end up building features for the wrong segment.

When Claap started selling seriously, the real buyer profile became clear.

Sales leaders. Revenue teams. People with budget, urgency, and clear problems.

That insight pushed the positioning, roadmap, and GTM into focus.

Robin’s recommendations for product people

If you are a founder or PM trying to build a product that actually delivers value, Robin suggests a few starting points.

• Read Shape Up from Basecamp
• Even if you never use their product, the way they think about cycles, scope and autonomy is extremely helpful for small teams going from zero to one.
• Follow Lenny Rachitsky
• His newsletter sits at the intersection of product, growth, and founder reality. Less theory, more lived experience.
• Use AI tools like Claude and Lobe as creative accelerators
• Not to build your full architecture, but to move faster from idea to first prototype and to help structure specs, tests, and design drafts.

Inside Claap, Robin even uses Claude as a first draft writer for product specs. He drops a few sentences about a feature, Claude does the research based on context and writes a full spec plus acceptance tests. Then Robin edits. The AI does the scaffolding, he does the thinking.

Where to follow Robin and Claap

Robin is most active on LinkedIn, sharing thoughts on product, AI and building in public. Claap itself is evolving fast, especially on the “conversation OS” side.

If you care about

• turning meetings into structured insight
• building remote first, async first teams
• or building products that start from real problems

you will probably see his name again.

And if you are still using three different tools for recording, summarizing, and searching through calls, you might want to give the product a try too.

‍