Open to new projects · remote · PL / EN
I automate business processes with AI agents, n8n and MCP.
I design and ship LLM-powered automations, AI agents and MCP servers, plus full-stack apps in Next.js and NestJS. I bring AI into places where data security and procedures matter — on-premise too, on your own infrastructure.
Ask my agent
It answers only from the content of this site.
01 — Services
How I can help
- 01
AI automation (n8n + LLMs)
For companies and teams losing time on repetitive work with emails, documents and spreadsheets.
What you get
- n8n workflows with LLM steps: classification, data extraction, summaries
- Integrations with email, calendar, CRM, spreadsheets and messengers (e.g. Telegram)
- Error handling, monitoring and hand-over documentation
n8nLLMwebhooksTelegram - 02
AI agents & MCP servers
For teams that want an AI assistant (Claude, Cursor, ChatGPT) to use their systems and data safely.
What you get
- An MCP server for your system or API — with dry-run mode, confirmations and validation before writes
- Agents that carry out tasks according to written procedures, with notifications
- RAG over company documents, including with locally hosted models
MCPagentsRAGQdrant - 03
Full-stack apps in Next.js / NestJS
For companies that need a dashboard, an API or an internal app — with or without AI features.
What you get
- Next.js front end (React, TypeScript, shadcn/ui), NestJS or Laravel API
- PostgreSQL, Docker, CI/CD; on-premise or cloud deployment
- Code in your repository, tests and a README
Next.jsNestJSPostgreSQLDocker - 04
AI training for institutions and teams
For public offices, NGOs and companies that want to use AI deliberately and safely.
What you get
- Hands-on workshops: AI assistants in everyday work, good prompts, model limitations
- Data security and AI usage policy in the organisation
- For IT teams: coding with AI agents, MCP, local models
workshopsonlineon-site
02 — Case studies
Projects and experience
Every write-up follows the same shape: problem → what I built → outcome → stack. Work projects are described anonymously.
Own projects
- 01ReadLiveOwn project · in daily use
An agent that works while I sleep
A personal AI agent system driven from Telegram: a daily job-offer digest, notifications from long-running work and custom agent skills.
Bun · TypeScript · Telegram Bot API · systemd timers · Python · Google OAuth · IMAP
- 02ReadLiveOwn project · the homelab as a lab
A private cloud for AI and automation
Own on-premise infrastructure: virtualisation, a TLS reverse proxy, external access without open ports, n8n, a vector database, monitoring and backups.
Proxmox VE · Proxmox Backup Server · Docker · Traefik · Cloudflare Tunnel · Tailscale · OPNsense
- 03ReadLiveOwn project · running in production in the homelab
A mail microservice for services and AI agents
One HTTP API for sending email from every service and agent: a separate key per service, a full delivery trail in PostgreSQL and a dashboard with GitHub sign-in.
NestJS · Next.js · TypeScript · Drizzle ORM · PostgreSQL · nodemailer · Auth.js
- 04ReadIn progressOwn project · in progress (API under construction, no UI yet)
From one sentence to a finished PDF — a small decision model instead of an LLM
Filling in a form from a single Polish sentence: a small local decision model answers closed questions, domain validation, a PDF from a DOCX template. Work in progress.
Python · FastAPI · Kev 0.8B · Next.js · Turborepo · Docker · pytest
Work experience (anonymised, code not public)
- 05ReadExperienceExperience · public sector · code not public
An MCP server with guardrails — an LLM writing safely to an internal system
An MCP server that lets an AI assistant log work in an internal tracking system: dry-run mode, two-step confirmations and validation before every write.
Python · MCP SDK · Keycloak (OIDC) · pydantic-settings · pytest
- 06ReadExperienceExperience · public sector · code not public
A RAG assistant over documents
A proof of concept of an assistant that answers questions from dozens of PDF documents: vector search, cloud and local models, streaming chat.
Next.js · NestJS · Turborepo · Qdrant · PostgreSQL · Claude · OpenAI
- 07ReadExperienceExperience · public sector · code not public
Talking to a database in plain language — streaming Text-to-SQL
A service that turns a natural-language question into an SQL query, runs it and streams the answer in a format compatible with the Vercel AI SDK.
Python · FastAPI · LangChain · Ollama · SSE · Docker · Traefik
Agent-ready
This portfolio is also an MCP server
Connect it to Claude, Cursor or any MCP client and ask your assistant about my projects. Read-only tools, generated from the same content as this site.
https://tomaszstec.me/api/mcpTools
- get_profile
- list_services
- list_case_studies
- get_case_study
- get_contact
03 — Process
How I work
- 1
Call
15–30 minutes: I learn the process, the data and the constraints, and tell you plainly whether AI makes sense here and where to start.
- 2
Demo / PoC
A small, working version on a real slice of your process — before you invest in the whole thing.
- 3
Delivery
Code in your repository, secrets kept out of code, monitoring, tests and documentation.
- 4
Training & support
Hand-over to your team, training, and — if you want — ongoing maintenance and development.
04 — Stack
Tools I use
- AI & agents
- ClaudeOpenAIVercel AI SDKMCPLangChainOllamaQdrantRAG
- Automation
- n8nTelegram Bot APIwebhookssystemd timersPython
- Backend
- NestJSNode.jsTypeScriptLaravelPHPFastAPIPostgreSQLDrizzle ORMRedis
- Frontend
- Next.jsReactTypeScriptTailwind CSSshadcn/ui
- DevOps & infrastructure
- DockerTraefikCloudflare TunnelProxmox VEJenkinsGitLabAnsiblePrometheusGrafanaKeycloak
05 — About
A developer who puts AI to practical use
I'm a full-stack developer. In my day job I'm a lead developer in the public sector. My main focus is growing AI-assisted development in the team: agents, skills, rules, MCP servers and project templates.
I'm also the technical lead of the application template the team starts new projects from, and of several systems — from requirements (PRDs) and diagrams to deployment. My stack: a pnpm + Turborepo monorepo, Laravel and PHP, Next.js, React and TypeScript, PostgreSQL, Docker, Traefik, Jenkins, Keycloak.
I pass AI working practices on to the team at daily meetings, through the template and documentation, and in training sessions.
I'm in the second year of an engineering degree in computer science, in the Programming with AI track, with an additional specialisation in artificial intelligence and machine learning.
My homelab is my lab: a virtualisation cluster, n8n, a vector database, monitoring and backups. That's where I try agents and automations before I recommend them to a client.
- Role
- Lead developer in the public sector
- In IT since
- 2018
- Studies
- Engineering degree, computer science — Programming with AI (year 2)
- Certificate
- IREB CPRE Foundation Level
- Location
- Silesia, Poland · remote
06 — Contact
Let's talk about your process
Briefly describe the process or the problem. I'll reply by email and suggest a short call.
I write in Polish and English.
Prefer to write directly?