Flutter
Unity
Infoshare
FlutterE2E
Aplikacja konferencyjna dla eventu IT — agenda, mówcy, networking i obsługa offline.
AI-Native DeveloperMobile DeveloperFullstack Developer
4 years of commercial mobile app development for Android and iOS.An independent, complete developer with my own agentic workflow — I deliver end-to-end.
Before the AI era, I built commercial mobile apps in Flutter and Unity. Today, with generative tools, I've added websites and web apps to that stack. I run delivery from idea to launch, adapting to whatever the project or client needs.
- Production code in Flutter (Dart) and Unity (C#)
- Claude Code: my own config, hooks, skills, subagents
- BSc in Computer ScienceTechnical English B2/C1
- Production code in Flutter (Dart) and Unity (C#)
- Claude Code: my own config, hooks, skills, subagents
- BSc in Computer ScienceTechnical English B2/C1
An AI workshop, backed by the 10xDevs certificate
I completed the cohort-based 10xDevs course led by Przemek Smyrdek and Marcin Czarkowski — an intensive program on software engineering with AI agents across the whole development cycle, from idea and plan through testing and deployment. The course was built around practice: every module closed with a concrete, working result built on a real project. The certificate, the four badges I earned, and the five areas the course covered are below.
10xDevs — cohort course
BRAVE / 10xDEVS · Przemek Smyrdek and Marcin Czarkowski
20 July 2026
Prework and 25 lessons across 5 modules, 6 live sessions (kick-off, 4× Live Q&A, celebration) plus weekly office hours. Final project: a custom AI agent and internal tooling built end-to-end.
Enlarge the 10xDevs certificate
- 10xBuilderFor delivering a complete MVP — from database and API, through the frontend and business logic, to tests and production deployment. The mandatory bar for completing the course.
- 10xArchitectFor designing architecture for larger systems and AI integration patterns at scale, together with modernizing existing code. A distinction for an extra challenge beyond the mandatory project.
- 10xChampionFor a fully working agent in CI/CD on GitHub, ready for teamwork: it reviews the diff on a pull request and posts a comment assessing the change. One of the cohort's extra challenges, beyond the mandatory scope.
- Best ProjectA distinction for the final project. Mine was a cross-platform app (Android, iOS, web) in a multi-repo setup, on a custom stack and with many original solutions.
Agentic Environment
Setting up the environment for working with AI agents: a custom toolset, project memory, and rules that keep the agent within the bounds of the task. This is where the agent gets the context and the frame in which it delivers a repeatable result.
10xDevs Workflow
A repeatable track from brief to working code: planning, review, and implementation as separate stages, with a clear split between what the agent does and where the human decides. On larger tasks several agents work in parallel, and every change is closed by a code review.
AI Development Quality & Maintenance
Quality and maintenance of AI-written code: a test plan, a custom QA agent, E2E tests, and bug reporting wired into a repeatable process. The standards that keep a project in good shape long after the first release.
Innovation with generative AI
Using generative AI to build new features and tools that genuinely speed up the work. Fresh techniques practiced right away on a working project, so they land straight in the product.
Large and legacy projects
Working with AI in large codebases and legacy projects: mapping the repository, scaling context, and modernizing legacy in the spirit of DDD. Introducing changes safely where the stakes are high, with a modernization plan and documentation generated along the way.
What I can deliver
First, what I have documented experience in. AI tooling and business processes — below, as areas I'm actively expanding into.
Mobile development — Flutter / Dart
- Android + iOS mobile apps from scratch and growing existing ones,
- architecture: Clean Architecture, Riverpod, Dependency Injection, SOLID, use-cases, Freezed, json_serializable,
- navigation and codegen: go_router, deep links, build_runner,
- data and backend: REST API, Firebase, offline-first with sync, Hive, shared_preferences,
- authentication: JWT with token refresh, secure storage, sessions, social login,
- device integrations: camera, QR / barcode scanning, Google Maps, PDF generation, POS / printers,
- notifications and monetization: FCM, in_app_purchase, RevenueCat, AdMob,
- UI and animations: responsive UI, implicit / Tween / Hero animations, Lottie,
- quality and delivery: CI/CD on Codemagic, flavors, Git-flow, code review, publishing to Google Play / App Store; testing as a standard part of the process.
- Flutter
- Dart
- Riverpod
- Freezed
- DI
- Clean Architecture
- go_router
- REST API
- Firebase
- Hive
- FCM
- RevenueCat
- AdMob
- Android
- iOS
- Google Play
- App Store
Apps in Unity / C#
- mobile and interactive apps built in Unity / C#,
- integrations: Firebase, REST API, Unity IAP,
- responsive UI (UI Toolkit / Canvas) and separating C# logic from MonoBehaviour,
- design patterns applied in practice and custom editor tooling that speeds up the team; focus on readable code and standards.
- Unity
- C#
- State Machine
- Object Pooling
- Scriptable Object
Common to both stacks
- code readability, architectural scalability, long-term maintainability,
- the full product lifecycle: planning → implementation → testing → release → maintenance,
- collaboration with developers, testers, project managers, clients.
Portfolio
An overview of my work: commercial mobile apps for Android and iOS, and my own projects built with AI agents.
Commercial portfolio —
As a Mobile Developer at Honeti — full lifecycle from planning through implementation, testing, and release to long-term maintenance.
- Flutter
- Dart
- Riverpod
- REST API
- Firebase
- Clean Architecture
- CI/CD
- Unity
Uprawnienia Budowlane
FlutterE2E
Aplikacja do nauki dla osób przygotowujących się do egzaminu na uprawnienia budowlane — testy, materiały, model subskrypcyjny dostępu do treści.
Gastro Ninja Klient
Unitymaintenance
Aplikacja klienta końcowego w ekosystemie Gastro Ninja — platforma do zamawiania jedzenia na wynos i dostawę, w modelu podobnym do pyszne.pl / pizzaportal.
Agentic programming — first-party deployments
AI is part of my production process. Four live projects where you can see it in practice.
Neatu Storage
- Flutter
- Firebase
- CI/CD
- E2E
A mobile app for managing home inventory: it tracks items by storage location (pantry, fridge, freezer), monitors stock levels, scans barcodes, and builds shopping lists from what's running low.
Neatu Dashboard
- Web
- E2E
The web companion of the Neatu ecosystem (neatu.app) — reach your inventory from the browser, alongside the mobile app.
Polylocale
- TypeScript
- AGPL-3.0
- Web
- E2E
My open-source localization tool. Format-faithful architecture, AI translation through the user's own API keys.
buzzards-soft.com
- Next.js
- React
- i18n
- E2E
The site you are reading, built with the active assistance of AI agents. The entire process is publicly documented.
How I work with AI
AI Native Development Workflow
The same controlled track from idea to deployment — with a clear split between what we do together, what the agent does on its own, and where a human decides.
Flow · idea → deployment
shared workmostly humanagent, unsupervised
- Idea · start
Context and plan
human + AITogether we settle the full scope of the change: the goal, the area in the code, dependencies, and the order of steps. Project context — conventions, decisions, architecture — is written down and fed to the agent. A concrete plan with clear scope boundaries emerges.
Plan approval
humanThe agent presents the finished plan and I go through it point by point — approving each step or correcting the scope. Code only starts after my approval.
- Agent at work · unsupervisedbuild ⇄ test ⇄ check · until everything passesBuild
The agent writes code and unit tests.
TestsThe agent adds integration tests.
Automated checkThe agent reviews the whole thing and writes up a list of remarks.
Assessment and decision
humanI read the collected remarks and the summary from the automated check, then decide: approve the work, order fixes, or reject it. The judgment of quality and fit to the plan stays with the human.
If the review catches something — we go back to "Context and plan" for new decisions.
Live testing
human + AII check the look and behavior on a real device — catching what only shows up in the finished product. Fixes and polish we do together, until everything works and looks the way it should ship.
Deployment
human + AIShared work built on CI/CD. The agent prepares the change description, saves it on a separate branch, and opens a pull request. An autonomous agent wired into the repository reviews the changes on its own. The final approval, merge, and deployment are done by me.
- Deployed
Under the hood
What stands behind this track — the methodical foundation and the concrete tools that keep the pace without losing control.
Spec-Driven Development
Nothing is built before a written, approved specification. The plan is the contract, and design decisions are recorded ahead of time — before any code exists.
Context-Driven Development
The agent works from durable, versioned project context — conventions, decisions, and memory from earlier sessions. Every task starts from the same, agreed foundation.
Specialized agent sessions
A separate session for planning, building, testing, and review — each with its own tools, prompt, and a single task to deliver.
Real tools in the stack
Live library documentation, checking the look in the browser, end-to-end tests on a mobile device. The agent works with real tools in my stack.
Repeatable processes as commands
Software-development steps wrapped into ready commands the whole team runs — from planning to reviewing changes.
Control over what lands in the repository
The agent works on a separate branch and opens a pull request at my request. A clear split: what the agent does, what I do, and how it ties into CI/CD.
About me
I'm a self-sufficient developer who runs projects end-to-end and has my own agent-based AI workflow. I've spent 4+ years building commercial mobile apps in Flutter and Unity, and generative tools have extended my stack to websites and web apps. I treat AI as a workshop I steer deliberately at every stage — from architecture through release and maintenance.
The core of my commercial experience is mobile apps. In Flutter I built products from the first line of code: REST API and Firebase integrations, authentication, data sync, offline mode, store releases and ongoing maintenance. In Unity I worked on interactive applications. In both stacks I keep the code readable and the architecture ready to grow years down the line.
The heart of my daily work is programming with AI agents. Claude Code is my daily driver, configured for specific projects: my own subagents, hooks, MCP servers and skills for repeatable tasks. I run the agent in a loop through review, exploring unfamiliar codebases, codemods and scaffolding, so I close more in the same amount of time. Control stays on my side — the agent doesn't merge on its own, and every output goes through build and my code review.
Working with agents shifts where the difficulty sits. Framing the task precisely, designing the context and verifying what comes back all matter more — because the model speeds up good decisions and bad ones alike. I treat it as engineering: I build verifiable loops where generation is fast and the quality gate stays hard. That lets me, as one person, run a project at a scope that used to take a team, and carry it from analysis through maintenance.
I'm a developer who deliberately directs AI agents — and because of that I deliver more quality, faster, across the whole project cycle.
I came to programming after fourteen years as an electrician and installer of CCTV, alarm, fire-safety and smart-home systems. On projects where software meets hardware — IoT, automation, smart buildings — that background gives me a genuinely unique perspective.
Education
Wyższa Szkoła Biznesu — National Louis University, Nowy Sącz
- Degree:
- BSc in Engineering
- Major:
- Computer Science
- Specialization:
- Business Application Programming
Zespół Szkół Elektryczno-Elektronicznych, Szczecin
- Degree:
- Technician
- Major:
- Electrician
- Specialization:
- Power Electronics
Additional qualifications
- SEP E + D
- Driving license cat. B
- Sailing license
Open to contracts, full-time roles, and interesting projects.
I'm the best fit where you're looking for a developer who already has working products under his belt and can work with AI in the workflow. If this profile fits what you're looking for — get in touch and let's talk through the details.