Sydney, Australia
Serge Kotlyarov
Grumpy CTO
I have spent my career building software, technology teams and companies. I am most useful when an ambitious idea needs clear judgement and a practical route to a working product.
My interests sit where engineering, science and business meet.
Pro-bono
Consulting
Humanity has lived through four industrial revolutions, driven by steam, electricity, semiconductors and the internet. AI is the fifth: neither the end of humanity nor a bubble. Drawing on my experience, I understand AI well and can help with:
- Prompting precision and clarity of intent
- Evaluation of results and quality judgement
- Multi-agent decomposition and delegation
- Failure pattern recognition
- Security design and compliance
- Context architecture
- Cost and token economics
- MCP servers; private, locally hosted LLMs
If you are building a business with AI and any of the above is not your focus or consideration, you may want to invite me for a coffee.
Selected work
Projects
A few ideas I am exploring.
BoletusOps
A corporate AI safeguard providing policy-based routing, audit trails, sensitive-prompt controls and cost governance - the controls your risk and compliance team is asking for. The project is currently on hold while I look for its first adopter. Would you like to be involved?
Mr PeaceDeal
I wrote my first browser game in 1998. It worked only in Netscape Navigator because Microsoft had its own “vision” for JavaScript. Today, a few prompts are enough for AI to build a game - from the engine to the graphics. For this experiment, I deliberately chose the gloriously trashy style of a 1990s arcade game. My only manual intervention was "slightly" correcting AI's sense of humour.
Notes and essays
Blog & publications
I write about AI, engineering, science, startups and politics. I plan to write more about the last one.
Grumpy CTOEssays and working notesA publications list will be added here.
Contact
Send me a message
Your details stay private and are used only to reply.