Vlad Shipilov

Vlad Shipilov

Founder Tobira.ai and Revoly.ai

About

👉 Building Tobira.ai, an open network where AI agents find clients, partners and deals for their humans. 👉 Also running Relovisa, helping founders with residency and citizenship strategies. 👉 Previously founded and exited Migrun (immigration tech), Product of the Day #2 on PH. 👉 Based in Porto. Here to launch products, support makers, and talk about AI agents and bootstrapping.

Badges

Buddy System
Buddy System
Gemologist
Gemologist

3

Top 5 Launch
Top 5 Launch
Tastemaker
Tastemaker
View all badges

Maker History

  • Tobira.ai
    Tobira.aiA network where AI agents find deals for their humans
  • MigRun
    MigRunResidence permits and visas in Europe 3-5X cheaper & faster
    Sep 2022
  • 🎉
    Joined Product HuntDecember 26th, 2021

Forums

Would ChatGPT or Claude buy from your company? We built a mystery shopper to find out

First, thank you all Tobira Site Agent took #1 Product of the Day on November 6, our second #1 this year. Thanks for every comment and feature request on Hypership day.

Right after the launch we built something new: a mystery shopper where the shopper is AI.

Up to 10 AI models (ChatGPT, Claude, Gemini, Perplexity and more) try to buy from your company, and we score every step out of 100:

  • Do they recommend you, or a competitor?

  • Can they find what they need on your site and get your prices right?

  • How much time and how many tokens do they burn?

  • And the big one: can they book a call or buy from you?

Lead quality and cold start - key differentiation questions

Great concept for AI-first lead generation. Key question - how do you validate lead quality within the agent network? If makers submit low-quality prospects and agents see bad signals, that breaks trust. Also curious about cold-start problem: new makers with no existing traffic need a bootstrap path into the network. How do you handle that?

Back after 5 months: what changed in how I build Tobira with AI agents

After our #1 launch in March I went quiet for almost five months, sorry for the silence.
When I came back, my old way of building with AI stopped working.

My Claude and Codex limits were gone in 2-3 days. Codex kept forgetting what the other sessions had done. For three weeks I couldn't ship one feature: a proper agent loop for our site agent, so it searches, checks the data and only then answers. I kept forwarding messages between the two AIs by hand, like a courier.

In the end I set up a harness: one shared space where every AI session, in any tool and on any computer, starts by knowing what we decided, what we tried and what's next. Within two days the stuck feature finally worked. Now my limits last the whole week.

I wanted to share this experience and discuss it here. The full story was too long for one post, so I put the detailed breakdown of how I build Tobira and how the harness took our product development to the next level on my Substack: https://founderbrain.substack.co...

View more