Frontend Engineer building AI-powered automation tools, enterprise UI systems, and mobile-first workflows.
I build practical AI-powered products and the engineering workflows around them - Angular/React frontends at enterprise scale, an Android AI automation app, and custom Claude Code / Codex agents that help me analyze, refactor, test, and document real software. AI as a multiplier, not a shortcut.
→ Automatist · Native Android AI workflow automation app
Workflow-first mobile automation built in Kotlin. Cloud AI provider profiles, scheduled execution via WorkManager, workflow import/export, Pro unlock through Google Play Billing, and on-device API-key handling. Local/offline AI is the next direction.
Custom skills, agents, and prompt packs for project-aware test generation, behavior-preserving refactors, migration planning, and large-codebase documentation. Built for real engineering work, not demos.
Document-management and ECM-style applications: shell/client architectures, iframe-based AI chatbot integrations, inbox and explorer UIs, and complex multi-step workflows designed for maintainability.
Reusable E2E architectures, page object models, fixtures, and CI reporting (JUnit, GitLab pipelines) that make large frontends safe to ship and fast to iterate on.
I'm not just experimenting with AI tools. I'm building a workflow where AI agents help analyze, generate, test, document, and refactor software - while engineering judgment, product thinking, and review stay firmly in control.
- AI-powered mobile automation - Automatist, on-device workflows, offline-model experiments.
- Enterprise frontend architecture - Angular/React shells, iframe integrations, ECM-style document apps.
- AI-assisted development - Claude Code, Codex, prompt packs, reusable engineering agents.
- Testing & delivery - Playwright + Vitest architectures, CI/CD reporting, accessibility-oriented E2E.
- Local-first & self-hosted AI - Ollama, Open WebUI, Docker Compose, Ubuntu Server, model routing.
Serious experiments, not random hobbies - each is a testbed for a pattern I want to use in real products.
- Automatist - mobile AI workflows on Android, with scheduled execution and provider profiles.
- Local AI stack - self-hosted Ollama + Open WebUI on Docker, privacy-aware routing experiments.
- Offline AI on Android - on-device model exploration for fully-local workflow execution.
- Claude Code prompt packs - reusable engineering agents for analysis, tests, refactors, docs.
- Browser & automation experiments - Chrome/Edge extensions, RSS-to-post pipelines, GitHub/GitLab automation, LinkedIn workflow tests.
- Self-hosted dev tooling - Linux + Docker stacks for local-first developer productivity.
| Frontend | Angular · React · TypeScript · JavaScript · HTML · CSS · Fluent UI |
| Backend & APIs | Node.js · Express · Python · Django · REST · PostgreSQL · MongoDB |
| Mobile | Kotlin · Android · WorkManager · Google Play Billing · Google Drive / OAuth |
| AI & Automation | Claude Code · Anthropic API · OpenAI / Codex · Ollama · Prompt engineering · AI agents |
| Testing & Quality | Playwright · Vitest · Jest · Page object models · E2E architecture · JUnit reporting |
| DevOps & Tooling | Docker · Linux · Git · GitHub Actions · GitLab CI · Postman · VS Code · JetBrains |
I ship features together with their tests, documentation, and review notes - not just code changes.
- Understand the real problem before touching code.
- Design before coding - architecture, data flow, failure modes.
- Use AI agents to accelerate generation, analysis, and refactors.
- Review and test every meaningful change; tests are part of the feature.
- Document decisions so future-me (or future-you) isn't guessing.
- Iterate - ship small, measure, improve.
On-device / offline LLMs on Android · model routing and provider-profile abstractions · Claude Code plugin and prompt-pack distribution · self-hosted AI gateways · local-first developer tooling.
