Afanasyevs — AI-native operating systems for people-run businesses

Your company already has software.

Accounting. CRM. ERP. Email. Spreadsheets. Maybe even ChatGPT.

People copy numbers into Excel.
They chase approvals on WhatsApp.
They forward emails.
They remember what happens next.
They call someone when a case does not fit.

Still manual work. Just digital.

What AI automation actually begins with

Another chatbot does not automate a company.

To take over operational work, AI has to know:

what needs to happen
which systems hold the information
which rules apply
what normal looks like
what counts as an exception
when a human decides

None of it is written down. It lives in systems, spreadsheets and people’s heads.

How it works

Built around what you already have.

Your accounting system stays. Your CRM stays. Your ERP stays. Nothing is migrated.

structured operational data
automation and orchestration
business rules
company knowledge
AI agents
escalation to a named person

Not another piece of software. An operating system for how the company works.

01
Build the brain

Record how the work is actually done. From live work, not interviews. Your systems stay.
02
Deploy agents

Move the routine to agents. Only what can be checked, undone, and has precedent behind it. The rest stays with your team, in writing.
03
Learn from exceptions

Route the rest to a named person. Their answer becomes precedent. The next such case closes on its own.

What changes

Routine runs without anyone pushing it.
People stop moving information between systems.
Excel stops being the company’s invisible operating system.
Specialists handle the cases that need them.
Unresolved cases go to the right person, and the decision is kept.
The next similar case needs less human involvement.

That is how the company brain develops.

Case study

Seven companies.
Seven ledgers.
One brain.

“This is our new personalised ERP.” — the group’s owner

An industrial minerals group across seven entities in Southeast Asia, Oceania and the EU. Seven ledgers synchronised in real time, one operating layer above them.

7
legal entities, one operating layer: 195
automation scenarios in production: 69
tables holding the group’s business logic: 6
months, fixed fee

Recognise your business?

Scaling business. Too complex for spreadsheets, too small for a Big Four ERP programme.
Analytics is hand-made. You need a live view of operations, but it has never been properly built.
A few irreplaceable people. Only long-tenured staff can close the gap between the written process and the real one.
No real owners. Things get done, but nobody truly owns the process.
AI on top of nothing. Assistants answer questions about documents; nothing answers questions about the business.
Automation tried and failed. You tried the IT approach. Now build it around how the business actually operates.

Start with a conversation.

Book a 45-minute call to discuss what keeps you up at night as a business owner.

Book a call hello@afanasyevs.com

Your people should not operate your software. Your AI operating system should.

Your company already has software.

Another chatbot does not automate a company.

Built around what you already have.

Build the brain

Deploy agents

Learn from exceptions

Seven companies.Seven ledgers.One brain.

Recognise your business?

Start with a conversation.

Your people should not operate your software.
Your AI operating system should.

Seven companies.
Seven ledgers.
One brain.