From a bank for entrepreneurs to an AI unicorn: an honest talk with Dmitry Salatov
No Bullsh!t, October 29, 2025, 73 min, 2,753 views. The guest is Dmitry Salatov, former founder of a mobile-first bank for entrepreneurs and now founder of an AI company building a canvas workspace for designers and creators that brings dozens of image and video models together in one interface.
| Published | |
|---|---|
| Length | 73 min |
| Views on YouTube | 2,753 |
| Language of the episode | Russian |
| Where to watch and listen | YouTube, Spotify, Telegram |
Date, length and view count according to data from the episode page on YouTube, as of the page update date shown in the footer.
In short
Dmitry traces the path from the bank, which he had to exit after 2022, through the search for a new niche, to the current product. The conversation covers the main failures of his career: intellectual property rights that were never properly secured, a conflict with a co-founder, and moving his family across two countries. From there come practical lessons on splitting equity, roles and responsibilities, aligning expectations between partners, and the question of focusing on one business versus running several in parallel. A separate block looks at where generative AI is going.
Key ideas
- The biggest failure at the bank was not securing intellectual property rights for the international structure at launch, which meant a full exit from the business after 2022 instead of keeping part of it.
- Conflicts between founders and investors almost always start when money arrives, so roles, shares and the worst-case scenario belong in documents from the start.
- Equity between founders should be calculated through contribution in money, time and social capital rather than split equally by default, and expectations have to be aligned explicitly rather than assumed.
- Focus on one business beats parallel projects, especially once there is a family: an advisor role is compatible with other work, an executive role at several startups at once is not.
- The product solves the fragmentation of AI tools: instead of dozens of subscriptions and chat windows, more than 30 image and video models live in one canvas workspace.
- The technical difficulty of an AI startup is the instability of other people's interfaces even at large providers, and dependence on external infrastructure where an outage hits your product directly.
The full transcript
The episode is recorded in Russian, and the full transcript of the conversation is on the Russian page of this episode. English and Spanish subtitles for the strongest episodes are rolling out on YouTube in waves.
Open the transcript