Available for selected projects

Full-StackDeveloper&AISystemsArchitect

Designing software, automation and AI systems that connect intelligent decision-making with real-world operations.

01 / About

I design systems, not just scripts.

I'm a technically focused developer and AI systems builder with a background in programming, Linux administration, scripting and technical design.

I combine software engineering with automation and AI to build systems that solve practical problems — from industrial workflows to intelligent software platforms.

Foundation

Technik Informatyk

  • Programming
  • Linux administration
  • Graphics
  • Scripting

Leadership

AI Startup — CEO

  • Strategy
  • Product
  • Business model
  • Finance
  • AI technologies
profile.ts
const fkoko = {
 role: "AI Systems Architect",
 focus: ["automation", "multi-agent"],
 approach: "systems over scripts",
 stack: ["TypeScript", "Python", "Linux"],
}

How the pieces stack

  • Businessstrategy · product · model
  • AI Systemsagents · orchestration · LLMs
  • Automationworkflows · scripting · VBA
  • Infrastructurelinux · systemd · servers
02 / Skills

A capability map, not a checklist.

Each area is a system I actually operate in. Expand a module to see the technologies underneath.

01

Programming & Scripting

Building software and glue code across languages and runtimes.

JavaScriptTypeScriptPython+3
02

Linux & Infrastructure

Running and deploying self-hosted systems on Linux.

LinuxsystemdServers+3
03

Automation

Removing repetitive work from real industrial pipelines.

VBASolidWorks automationIndustrial workflows+1
04

AI & Multi-Agent Systems

Designing controlled autonomy around language models.

LLMsAI agentsMulti-agent architectures+3
05

Product & Business

Connecting technical architecture to business outcomes.

Product strategyStartup developmentBusiness modeling+2
03 / Work

Selected Projects

Systems, automation and AI infrastructure built around real-world problems.

01
Active

AI Consulting Startup

Co-founder & CEO

AI Navigators

Visit

AI consulting startup focused on helping small and medium-sized businesses in Poland adopt automation and AI-driven workflows.

AIAutomationConsultingStrategy
02
In development

AI Sports Analytics

Kineo V2

Training analytics and AI coaching platform designed for competitive athletes, built around a deterministic decision engine that feeds an AI coaching layer.

AILLMAnalyticsDecision Engine
03
Prototype

AI Infrastructure

Nexus Director / Agent OS

Autonomous multi-agent orchestration system designed to coordinate specialized AI agents through controlled, deterministic workflows.

systemdOrchestrationNemotronDeepSeek
04
Live

Product & Web Application

PC Configurator

Visit

Interactive PC configuration tool that helps users assemble compatible components and plan a complete computer build.

Web AppPC BuildingUXProduct

04 / Case Study

Nexus Director / Agent OS

A controlled execution environment for multi-agent workflows, where autonomy is combined with deterministic orchestration.

The Problem

Traditional agent systems often let multiple agents act without sufficient control. Autonomy without boundaries introduces real operational risk.

  • Unpredictable behavior
  • No control over flow
  • Difficult debugging
  • No deterministic guardrails
NEXUS DIRECTOR
ORCHESTRATORACTIVE
DECISION LAYERREADY
CONTROL GATES04
AGENTS07
MODEL ROUTERONLINE

The Approach

The system is designed as a layered architecture. Every request passes through deterministic checkpoints before and after any agent or model touches it — so control never leaves the system.

The language model is one component inside the flow, not the decision-maker for the whole flow.

Agents do not operate without control.

Architecture Principles

01

Controlled autonomy

Agents can act autonomously, but only within explicitly defined constraints and boundaries.

02

Deterministic control

Critical decisions are not left to the language model alone — a deterministic layer owns them.

03

Process isolation

Agents run as controlled, isolated processes managed through systemd rather than free-running loops.

04

Model abstraction

The system integrates different models without coupling the whole architecture to a single provider.

Result

A controlled execution environment for multi-agent workflows where autonomy is combined with deterministic orchestration and explicit validation layers.

05 / Contact

Let's build something useful.

Have a technical problem, automation challenge or AI system in mind? Let's talk.

Email
LinkedIn
GitHub