I'm a Technical Product Manager based in Warsaw, focused on CRM, SaaS, data platforms, and AI-enabled workflows. At Procter & Gamble, I work across CRM technology, media optimization, and advanced analytics, turning complex systems into practical product outcomes.
I like building where product, data, and automation meet: MarTech platforms, growth systems, analytics pipelines, and personal AI agents that actually get used.
- Building and shipping AI-assisted products with a stronger bias toward finished work.
- Moving personal automation from cloud-hosted experiments to local, self-hosted AI infrastructure.
- Applying my MSc in Big Data and Advanced Analytics from the Warsaw School of Economics to practical product and automation problems.
- Exploring SaaS monetization, data platform architecture, and agentic product workflows.
- Getting more hands-on with code, closer to implementation details, and stronger on the technical foundations needed to push agentic frameworks forward.
Over the past year, I've been building personal AI agents and automation systems, moving from cloud-hosted bots to local, GPU-accelerated multi-agent workflows.
Clawdbot - WhatsApp AI assistant
A Node.js WhatsApp gateway that evolved into a multi-agent orchestration layer. It is migrating from EC2 toward a local Yahboom Jetson Orin Nano Super 8GB setup as a self-hosted AI compute node.
Node.js Ollama PM2 Tailscale Cloudflare Tunnel OpenAI Proxy
Multi-agent architecture
An 11-agent system adapted from the My-Brain-Is-Full-Crew framework, combining reused agents with custom devops, Clawdbot manager, and researcher agents. Tasks are routed across local Ollama models, OpenAI access, and Anthropic API when harder reasoning is needed.
Other experiments
- Polymarket betting agent: Python, Docker Compose, and
py-clob-client, starting with paper trading. - Autonomous coding loops: Claude Code CLI workflows controlled from mobile through SSH and tmux.
- FotoSeller: AI product photography pipeline for MercadoLibre and Amazon sellers using Gemini, Ideogram, and Claude Vision.
- Portal Trax: First Android app experiment, built while leveraging Claude skills to learn mobile foundations and ship a practical workflow.
- Product and growth: Roadmaps, adoption, engagement, experimentation, and global stakeholder alignment.
- Data and analytics: SQL, BigQuery, GCP, Power BI, APIs, dashboards, and insight-to-action workflows.
- MarTech and CRM: Braze, Segment, lifecycle messaging, audience strategy, and platform operations.
- Ways of working: Agile, Scrum, Kanban, and practical delivery habits for cross-functional teams.
- Hands-on product development with AI and agentic frameworks.
- Workflow automation, AI agents, and practical developer tools.
- SaaS, MarTech, CRM, and growth products where product thinking meets implementation.
- Projects where I can get closer to code, learn the foundations, contribute clearly, and help ship.
- Product strategy and analytics.
- CRM, growth, and SaaS.
- Personal AI agents and local LLM workflows.
- Career pivots, resilience, and building momentum after hard resets.
Brain tumor survivor.