AI system architecture
LLM, OCR, document-heavy automation, integrations, and data-intensive products with explicit components, boundaries, constraints, and implementation paths.
AI Architecture / MLOps / Production Agents
AI systems architecture, MLOps, and production-grade agent workflows.
I work on the engineering layer around AI systems: architecture, LLM workflows, agents, evaluation, human-in-the-loop control, Kubernetes delivery, CI/CD, observability, and cloud infrastructure.
Prototype → production
Clear path through system boundaries, rollout discipline, observability, and iteration—not just notebooks and POCs.
AI agents
Human-in-the-loop workflows with tools, escalation, and audit-friendly behavior.
Cloud & MLOps
Production deployment pipelines, monitoring, lifecycle discipline, CI/CD alignment.
Engagements
AI platform, MLOps, and cloud engineering work in Polish and English, with emphasis on systems that can be operated, reviewed, and improved.
Profile
Architecture first, production always.
AI Systems Architect / MLOps Engineer focused on architecture, AI agents, model operations, and platform engineering. I work across solution design, Kubernetes-based deployments, GitOps-driven CI/CD, cloud-native infrastructure, operational reliability, and security. Long-term, my direction is reliable AI systems for high-stakes domains such as space, defense, and critical operations, where systems thinking and operational discipline matter.
Current focus AI platform engineering, LLM systems, production agents, MLOps, cloud infrastructure, and reliability work around systems that need more than a prototype.
Work
Engineering problems I tend to work on.
A practical view of the architecture, platform, and operational topics that appear across my AI and DevOps work:
- Turning LLM prototypes into maintainable systems with clear architecture and operational boundaries.
- Designing workflows that handle real documents, imperfect data, and repeatable review paths.
- Building agents with human approval, auditability, tool use, and escalation instead of opaque automation.
- Adding MLOps, monitoring, evaluation, and deployment discipline around AI features.
- Assessing technical feasibility before implementation choices become expensive to reverse.
Focus areas
Where architecture meets delivery and operations.
The common thread is production-oriented AI: not only model behavior, but also the platform, delivery process, reliability practices, and feedback loops around it.
AI agents and workflows
Stateful workflows with tools, retrieval, evaluation, escalation, and human approval so behavior can be inspected, constrained, and improved.
MLOps & platform engineering
Deployment, CI/CD, monitoring, experiment and model lifecycle, plus cloud-native infrastructure aligned with repeatable software delivery.
Reliability and production operations
Kubernetes, GitOps, Terraform, observability, incident handling, and runbook-oriented operational work around AI and platform systems.
Core skills
Architecture, platform engineering, evaluation, delivery.
AI systems architectureLLM workflowsAI agentsRAGMLOpsEvaluationKubernetesGitOpsDatabricksTerraformAzure AILangChain / LangGraph
Experience
Selected roles and projects.
A focused view of work most relevant to AI systems, MLOps, cloud, automation, and architecture.
Craftware
AI Platform Engineer
AI platform engineering
Designing and building AI platform capabilities for automated technical and engineering documentation, with production MLOps, Kubernetes, GitOps, CI/CD, and stakeholder feedback loops.
Holisticon Poland
MLOps Engineer
MLOps and AI consulting
Working on ML platforms with AWS, Databricks, MLflow, Unity Catalog, Terraform, model lifecycle practices, production support, and selected AI agent automation.
Inetum / Capgemini
Cloud DevOps Engineer
Cloud and DevOps architecture
Cloud infrastructure migration, Terraform-based environments, GitOps with Helm and FluxCD, CI/CD, Docker, Kubernetes, logging, and automation in close cooperation with architects.
AIOps / DevOps projects
Junior AIOps & DevOps Engineer
Automation and applied AI
AI image classification pipelines, OCR-based document extraction, internal process automation, Jira and Confluence automation, and Python/Bash scripting for operational workflows.
Certifications
Credentials that support the consulting profile.
AWS Certified Machine Learning - SpecialtyAWS Certified Solutions Architect - AssociateMicrosoft Certified: Azure AI Engineer AssociateMicrosoft Certified: Azure Data Scientist AssociateHashiCorp Certified: Terraform AssociateISO 27001 Internal AuditorCybersecurity Representative: ISO 27001, ISO 22301, GDPRStanford Online XCS234 - Reinforcement Learning
Education
Computer science plus applied AI, cloud, and security studies.
Master's degree in Computer Science. Postgraduate studies across AI project management, cybersecurity, Cloud DevOps, and cloud solution architecture.
Publication
Co-authored SME implementation playbook — architectural depth up front.
Co-author of “Gotowy plan wdrożenia systemu AI w MŚP. Praktyczna dokumentacja prawna, informatyczna i ocena biznesowa projektu”, aimed at pragmatic AI rollout for SMEs across legal framing, ICT documentation, governance, and business judgement. My authored contribution focuses on technical system architecture and AI agent design so the organisational advice sits on workable engineering footing.
View publicationTalks
Speaking on practical AI implementation.
Akademia Leona Koźmińskiego / May 2026
AI w MŚP - od pomysłu do wdrożenia
Practical talk on designing AI agents with state, tools, retrieval, human approval, and production safeguards — not just chatbot demos.
Contact
AI architecture, platform engineering, and MLOps work.
For relevant engineering, architecture, consulting, or collaboration topics, send a short note with the context and constraints.