# The moment Digiotouch AI started

Founder's Note by Soumya Kanti Datta, 2026-11-12

I still remember the exact moment Digiotouch AI started - though I had no idea at the time that it was a beginning at all.

It was late 2022. We started working on **EU-funded** [ZeroW](https://www.linkedin.com/company/zerow-project) project, and [Debolina Paul](https://www.linkedin.com/in/debolinapaul) and [Maria Aiello](https://www.linkedin.com/in/mariaiello), had taken on something genuinely hard: understanding 9 different pilots end-to-end, scoping their sustainability goals, and building individual assessment road map for each one.

During our weekly team meetings, I'd listen to how they actually got the work done. They'd record the Teams call. Download it. Then re-watch the whole thing - often more than an hour - pausing, scrubbing back, taking manual notes. **And only after all of that could they begin the real work**. Same pilot, several meetings. Then again, across all pilots. 😮💨

I kept thinking about it after those calls. These were two of the sharpest people in my team, and they were **losing hours just re-watching videos to find answers**. It bothered me a lot. So I started asking around and quietly, almost everyone admitted they did the same thing. With remote and hybrid work clearly becoming the norm, I knew this wasn't going away.

So at the start of 2023, I just started building something. It was mainly for us at [Digiotouch R&D](https://www.linkedin.com/company/digiotouch-rtd), internally, at first. But **that became Digiotouch AI** as it took shape.

**The thing I cared about the most is this: a meeting is never just words**. People share slides, drop a PDF, walk through a design doc, debug software together, brainstorm together and **they rarely read any of it out loud. So much of the meaning is in what's shown, not said. I wanted a tool that understood that - one that watches the whole video, the visuals and documents included, and lets you simply ask for what you need**.

Getting there wasn't smooth. 😅 My PoC worked, but it was quietly burning $10-15 every single day. Impressive in a demo, completely unviable as a product. It stayed that way until [Marharyta Domnich](https://www.linkedin.com/in/marharyta-dekret) took over, rebuilt parts of it from scratch, and turned a clever experiment into something we at [Digiotouch OU](https://www.linkedin.com/company/digiotouch-ai) could actually stand behind.

And the result is a tool you can talk to like this:

"Explain me the figure on the slide 3 Maria put up about pilot 3?" "There was a diagram in the design doc someone shared - what did the architecture use for data ingestion?" "Which slide had the sustainability roadmap on it?" Upload a video recording, or capture your meeting live with Digiotouch AI agent or desktop apps, and just ask.

**It began as our own problem. I'd love for it to solve yours too**. 🚀

[**Sign up today**](/en/sign-up)