Looking for people building for creators. Let's connect!
I'm Andrew, based in London, founder and CEO of Kineto, with over a decade in content platforms, music products, and voice assistants. Most recently, before Kineto, I created and led Junie, the JetBrains coding agent, got to scale it from zero, watch it actually start shaping how teams work, and learn what scaling an agent really takes.
Now I'm taking that experience into a different industry: the creator and influencer economy.
Two things I'd genuinely love to talk to people here about:
The creator economy itself. Tools, platforms, business models, all moving fast. I'm a fan of the take that we can't really automate creativity, but there are 20 things around it that agents can genuinely help with. Not using AI is not on the table, that ship sailed. The interesting question is how we build products that take the boring 20 and protect the part people actually got into this for. Always curious to meet others working in this space: image, video, audio, voice, anything adjacent.
The AI adoption gap. Powerful tools, a public mostly tired or skeptical of them. The product question that interests me most isn't "how powerful can we make this," it's "how do we package it so a normal person actually wants to use it?" Best products feel embarrassingly simple. Most of today's AI feels the opposite.
If either of these is your thing, say hi.
Replies
Well hi.
I've come to a personal conclusion that AI on its own, in isolation, isn't that useful. I genuinely believe that when packaged correctly — giving the user not AI itself, but rather themselves with AI as a supporting function — it delivers far more value. That's exactly how I'm building Sidenum. AI as a tool, not the product.
Nice to meet u.)
Hey @mykyta_symonenko , nice to meet you!
AI as a tool, not the product, that's good framing framing. What important is the user with their goal, AI is just what makes the gap shorter. Curious how how you handle in Sidenum the personalization without making the user "configure" it too much, have you already found a good balance?
@andrewzakonov That's exactly the challenge we focused on from day one.
Our core belief: the user should never have to "configure" anything. The AI does it for them — invisibly.
Phase 1: Frictionless Setup The user states their goal or drops a job posting link. Then a guided survey establishes the full context: country, experience level, specialization (e.g. Growth PM vs Technical PM), industry, existing tools and skills, and one open-ended question. No technical terms, no complex setup. Just answers.
Phase 2: Adaptive Learning Each block has two modes — learning and exam. The system remembers everything. Failed an exam? The next simulation already knows exactly where you struggled and focuses there — not just by adjusting difficulty, but by actively guiding you back to that specific gap.
Working with a tool like Roadmunk? The system tracks what you do, where you make mistakes, and suggests improvements in real time. That interaction becomes context.
Phase 3: It all connects Here's where it gets interesting. If you learned Roadmunk during an earlier block, and now you're in a roadmap-building simulation — the system uses that. It helps you build a roadmap inside Roadmunk, adapted to your specific goal. Want to launch a coffee shop? Your roadmap reflects that.
The user configures the system through their actions, not through toggles and settings.
Would love to hear your opinion!
I also posted about this in (p/introduce-yourself). It is right after your post)
Stylar
@mykyta_symonenko Well said