Operator · AI-native operating models

Designing AI-native companies, where people manage agents, not routine.

9+ years in e-commerce and operations. Currently building systems in which hidden team knowledge becomes managed AI processes, and growth comes from role multiplication, not from headcount inflation.

Skolkovo AI Shift & Practicum #29  ·  Multiple AI products in production  ·  E-commerce & global marketplaces
Category

Most companies bolt AI onto a cart. I’m building the car.

Over the last two years, every mid-sized company added “AI chats” and “GPT helpers”. Accountants use ChatGPT for emails, marketers — for copy, operations — for reports. That’s an AI-augmented company. A lot of noise, the operational structure hasn’t changed.

An AI-native company is built differently. Processes are designed from the start around human-agent collaboration. The employee doesn’t use a neural net to help themselves — they design a team of agents and manage it. These are different categories, and they shouldn’t be mixed.

Old world
AI-augmented
  • GPT bolted onto old processes
  • Growth through hiring
  • Employee = routine operator
  • Knowledge in key people’s heads
  • Bus factor grows with the company
New world
AI-native
  • Processes designed for agents
  • Growth through role multiplication
  • Employee = agent architect
  • Knowledge in a managed system
  • Bus factor solved architecturally

The difference is not in the number of tools — it’s in the operating model. Companies that complete this transition first will hold the advantage of the next decade. The rest will stay with “AI pilots” and wonder why growth doesn’t come.

Proof

What stands behind these words

9+ years
Operational work in e-commerce and global marketplaces. Not theory from articles.
Production
Several proprietary AI products running in production, serving real operations every day.
Skolkovo
AI Shift program (4 months) and Practicum #29. Faculty and network — first-hand.
10:1
In specific processes inside my companies, one person closes work that used to require a team.

I don’t sell “AI adoption”. I build operating models I use myself. Every claim in this text comes from what already works, not from what sounds good.

Manifesto

Three texts that make it easiest to see how I view the world

If you want to understand before writing — start with the first. The rest will make more sense after.

Subscribe to the Telegram channel New notes on AI-native operations and breakdowns from practice
About

Aleksey Rodkin

Entrepreneur and operator. Nine years building companies in e-commerce and working on global marketplaces. Studied at Skolkovo — AI Shift program and Practicum #29.

Currently shifting focus to AI-native operating models: systems where employees don’t close routine manually, but design and control agents instead. I’m interested not in “AI tricks”, but in how companies reduce bus factor, extract hidden knowledge and turn processes into managed systems.

Fewer people — more results. Less coordination — higher speed. That’s what an AI-native operating model is.

Contact

If there’s something to discuss

If you’re writing about something real — tell me briefly:

  1. Who you are and what you’re building
  2. Why now
  3. What you want to discuss: partnership, joint project, talk, strong implementation case
  4. A link to your product or profile

That’s how we’ll get to the point faster. Generic proposals without context — pass.