Karol Rogowski
Karol Rogowski · Warsaw

AI Product Builder
for companies that move fast.

I design, build and ship AI automations for small companies - end to end, in weeks, not quarters.

Get in touch
available · Q3 2026
Location
Warsaw, PL · CET
Response time
Within 24h
Engagement
Solo · 1–2 projects
• Examples

What can I automate in your company with AI?

// another case? get in touch →
01

Google Ads & Meta reporting

An agent pulls data on a schedule and writes the takeaways - the report lands in Slack, email or Notion.

02

Meeting summaries

Decisions and action items written up minutes after the meeting - into your CRM, email or knowledge base.

03

AI assistant with your knowledge base

It knows your documents, processes and clients. Ask in chat, get an answer with full company context.

04

Client follow-ups

Automated, context-aware messages at the right moment. Reminders for overdue invoices or projects awaiting confirmation.

05

Support ticket handling (help-desk)

AI classifies and answers repetitive, simple questions. The rest it escalates to a human with a ready summary and suggested solutions.

06

Sales & lead handling support

A lead comes in, AI qualifies it, enriches the data and tells the rep what to do next and when.

• About

I build, I ship.

Most AI projects stop at a demo. Mine reach production. First an interview and process mapping, to know what's actually needed. I build, we validate, we test together and ship to production. At the end I train your team, so you can grow what we built on your own.

I'm a freelancer. I take one, at most two projects at a time and work closely with the founder and the users. If that's how you want to work together - we'll probably get along.

Stack: Claude · Claude Code · Supabase · n8n · Vercel · Replit · Posthog
Book a consultation 30 min · free
• What I build

Two things, done properly.

01

AI automations

I build automations powered by AI agents that handle specific tasks at specific times - sales, operations, support or reporting. Not rules in Make, but tailored scenarios with agents that know what to do, even when the situation changes.

Why an agent, not a plain automation Collapse
  • Handles exceptions - a plain automation breaks when something is off. An agent understands context and knows what to do next.
  • Doesn't need perfect data - rules in Make need exact formats and triggers. An agent reads intent, not structure.
  • Improves from mistakes - a classic automation always does the same thing. An agent can be fixed with a single instruction, without rewriting the flow.
02

AI implementations

I build the AI-native foundation of your company: a knowledge base, a second brain, processes built into AI. I run the interview, map, build and train the team. You walk away with a company where AI executes and humans oversee.

What being AI-native gives you Collapse
  • Company memory that never forgets - decisions, processes, projects - all available in a second through AI that understands the question, not by digging through folders.
  • Onboarding in days, not months - a new hire asks AI instead of interrupting you. Instant answers, consistent with how the company actually works.
  • Knowledge doesn't leave with people - when someone leaves, their context stays in the system. The second brain belongs to the company, not the person.
• Contact

Got something to build?

Drop me a line or book a consultation. I reply within a day.

Book a consultation 30 min