Cloud / Platform / AgentEngineering & Operations
I enjoy helping agents manage dramas of computing machines
Since 2011, I have focused on removing human labour from the deployment, operation, and monitoring of software across cloud and on-premises infrastructure.
Over the years I have worked in a range of technical and business roles within enterprise, startup, and public-sector organisations. The emergence of AI agents has brought me the great joy of harnessing efficiency gains while mitigating the risks of nondeterminism. I am now looking for opportunities where I can leverage my experience to develop agent-driven approaches to delivering and operating platforms in the cloud.
I am customer and delivery value oriented. I am happy to help you with platform delivery, team development and technical business infrastructure. I enjoy working as a hands-on engineer, team lead or in product development. I can bring my own approach or adapt and improve your strategy. I have learned and forgotten many technologies on my journey, but I am always open to finding and learning the best approach to the challenges ahead.
Build and operate platforms from initial architecture and prototyping, through delivery of automation, observability and reliability, to resilience, compliance and cost optimisation.
This includes managing vendors, roadmaps, deliverables, estimations, presentations and all the work related to the delivery.
Develop and nurture an open, value-driven engineering culture striving for autonomy, ownership and collaboration as a foundation for an excellent delivery-oriented team.
I am used to working with cross-site teams across timezones stretching from California to India. I mitigate the silo mindset by establishing joint ventures with different teams to align on a common set of goals serving the customers.
Delivery naturally feeds understanding of the target consumer community, power users and stakeholders. User experience problems, missing capabilities and recurring issues create an opportunity to focus on valuable deliverables.
I act on these findings, assess the wider impact of the new features and develop a user-facing roadmap to continuously improve the value of the platform to users and stakeholders.
Enterprise scale, startup velocity and public-service responsibility.
Waypoint was my ticket to experience agent-driven product delivery first-hand. I took over the CTO’s Cloud & DevOps responsibilities and gradually embraced an agent-driven approach to all aspects of my work.
While providing evidence for the SOC 2 Audit I was exploring ways to automatically align our compliance via Vanta MCP just by developing merge request skills. I established Docker Sandbox to restrict agents’ access to cloud credentials and as one of the agent sandboxes for the product’s dynamic pipeline capability. I scoped permissions for agents to perform routine root-cause analysis in the cloud, performed agent-driven migration of production environments between hyperscalers and code bases. I used agents in E2B platform delivery, Terraform development and all other aspects of continuous delivery.
I was hired as Big Data Platform Engineer for Hadoop deployments in AWS. I started to work on an in-house, AWS-native Machine Learning Platform and within the first year I became delivery lead of a Data Science platform powered by Dataiku. I stood up a platform with my team from nothing to an established enterprise platform providing Data Science, Data Processing, Machine Learning and AI-driven capabilities. By continuous improvement and efficient delivery, we spearheaded emerging enterprise enablers such as MCS 2.0 AWS Accounts, Kubernetes and enterprise-aligned Terraform Modules. We delivered Karpenter-powered scheduling and drove integration with the Enterprise Agent Platform.
As Product Line Architect I was responsible for technical evaluation and delivery of new capabilities. I delivered Code Ocean and Domino Data Science Platform for evaluation by Research Laboratories. Once Domino was chosen to power one of the analytical components in a new Clinical Pipeline, I worked closely with MRL & MRL IT to deliver a Business Critical System with my team.
In collaboration with AI Compliance and DevSecOps teams, we provided a Hugging Face model ingestion gateway, enabling scanning of models and storing them in Artifactory. We exposed onboarded models to the community via an Enterprise AI asset catalog. Meanwhile I was driving engagement with Enterprise Architecture to integrate our product line with the emerging Agentic Platform.
My responsibility was maintaining and improving operations, documentation, deployments, monitoring and security of services deployed across different infrastructure providers. Angelcam operates Cloud Recording, Streaming and Broadcasting services integrated with IoT sensors to provide a real-time security service for customers via mobile and web-based interfaces. I ran our operations across AWS, Digital Ocean and multiple bare-metal providers.
Among the initial deliverables was containerisation of Rust and Python-based services, their continuous delivery and autoscaling in Docker Swarm. I moved observability and alerting to Prometheus and exposed metrics to the team with Grafana. A big part of the job was building trust with customers, as it depended on reliable connectivity to AWS. In the security market, each downtime in recording might be a reason for cancelling a subscription. I made sure uptime of different services was not just high but also transparent and publicly observable.
My job was to deliver and continuously improve web archiving services for the National Library. Early in my career I had the opportunity to participate in a staff exchange at the Internet Archive in San Francisco. Among other things, this helped me understand configuration implications and behaviour of their archival crawler and kick-started my knowledge about how huge amounts of data are stored and processed in a meaningful way.
I designed and implemented recurrent Czech domain and topical crawls. I maintained a couple of on-premises storage servers with new hard drives, including a couple of RAID 6 resurrections. I was operating an archival retrieval service, our Web Archive websites and other services, research projects and prototypes in vSphere or on bare metal. Before I left I took responsibility for the Application Services Department, as I wanted to drive technical collaboration inside the library and across Czech and foreign libraries.
There is always more work to do.
My studying of human civilisation and playing games formed my ideas on approaching the Civilisation genre. With recent improvements in LLMs and agentic tools I was able to deliver prototype components beyond my programming, academic and mathematical skills. I embrace a dark factory approach and focus on iterative specification of game systems and developing visual appeal.
To help agents resolve technical issues, I usually instruct them to conceptualise simulation and computations for me. This greatly simplified navigating technical decisions affecting performance and behaviour of the simulation. Driving artistic aspects of the game, such as music and art, is way more tricky, but I am slowly getting better at adopting what generative AI can do well.
This project helps me greatly in assessing capabilities of frontier models and their quirks. I found it very efficient to dedicate a server to the agents and interact with them remotely. Without the recent rise of AI capabilities I would never be able to make any meaningful progress as a salary man.
Before my daughter was born, I used my free time for contracting. My clients were my ex-employers and I worked on a project basis.
CHARLES UNIVERSITY
B.A. Information Science & Librarianship
New Media Studies / Discontinued before thesis
MASARYK UNIVERSITY
Audiovisual Culture / Discontinued before thesis
HACETTEPE UNIVERSITY
Information Management / Erasmus
Czech / Native
English / Full professional proficiency
French / Elementary proficiency
RECENT LITERATURE & PODCASTS
Ancient and Prehistoric Economies
History of Technology
Contemporary Society and Conflicts
Supply chain, AI & Macroeconomics
Music, Art & Movies
Old, New & Weird
Recent Games
Elden Ring, Death Stranding, Cyberpunk 2077, Civilisation 7, Humankind, Disco Elysium
Prague / Czech Republic
Cloud, platforms, agents — and the people behind them.