Point of View
Automation works when PM, data, and operations use the same truth.
My approach starts with constraints in the real workflow, then maps AI and automation to remove those constraints without sacrificing governance.
-
PM becomes a system supervisor, not a status relay.
When agents can update project tools against explicit Definition of Done rules, PMs move from manual sync to quality control and stakeholder alignment.
-
Process truth must come before automation scripts.
Process mining and flow diagnostics identify where work stalls, so the first automation targets bottlenecks instead of creating new busywork.
-
FinOps and reliability are product responsibilities.
Payment settlements, POS integrations, and incident telemetry are not side concerns; they are the operating core for multi-site business delivery.
Selected Proof
Projects where I moved teams from manual operations to supervised automation.
-
Production Wrike MCP for AI-assisted task orchestration
Introduced AI-driven PM updates and blocker signaling with governance rules inside a live PM environment.
-
Settlement reconciliation between POS activity and banking reality
Built real-time visibility to detect payment mismatches early and improve operational confidence across branches.
-
Document tooling with anonymization for enterprise-safe LLM usage
Enabled faster processing of operational documents while reducing privacy exposure before cloud AI processing.
Writing
How PM teams get better with AI and process insight.
Credential
Validated training in agent systems and practical GenAI tooling.
Contact
Open to Head of Automations conversations.
If you are hiring for automation leadership, AI-augmented PMO, or process-driven ops transformation, send a note.
Warsaw · Remote-first · Polish & English