Independent consultant based in London. I join teams as a technical lead, architect, or escalation engineer — usually on problems involving Kubernetes, platform delivery, or putting LLM agents into production without embarrassing the company a year later.
Engagements across UK government, defence and national security, finance, critical national infrastructure, regulated public sector bodies, and startups that survived Series A. Clearance held as required.
Cluster upgrades, capacity planning, PDBs, HPA tuning, the slow unglamorous work of keeping on-call short.
Right-sizing estates, migration sequencing, cost models that survive contact with a finance director.
Internal developer platforms that remove friction without removing guardrails. Measured in deploy frequency, not slide count.
Policies devs actually read. Toolchains that fail open on false positives, closed on real threats.
SLOs tied to user-pain, not CPU. Error budgets as a decision tool, not a vanity metric.
Agents, MCP servers, retrieval pipelines. The operational plumbing underneath — evals, cost attribution, failure modes.
Auto-generated, type-safe Telegram Bot API client for Go. Tracks the upstream Bot API v10.0 spec — 176 methods, 301 types, 1428 generated tests — so the library never lags behind Telegram releases. Batteries-included: long-polling, retries, context cancellation, structured errors. MIT licensed, zero hand-written API surface, fully tested via codegen.
Patent-pending vector database that stores no vectors. Instead of persisting embeddings, GraphANN navigates a graph over compressed source text and recomputes embeddings on demand for the handful of nodes it actually visits — typically 200–500 per query. Result: ~97% less storage than a conventional vector DB, recall on par with a flat index, and hot-swappable embedding models with zero re-indexing. Ships as a single zero-dependency binary with multi-tenant isolation, zone-aware HA clustering, and pluggable inference backends (Ollama, ONNX, OpenAI, custom). A storage-efficient replacement for the usual vector database tax at production scale.
Developer productivity tool that transforms GitHub activity into a gamified experience. Analyses contributions and generates interactive dashboards with 115+ achievements across 26 categories. Features team analytics, GitHub Action integration, smart caching, and bot filtering.
Persistent memory system for Claude Code that enables long-term context retention across coding sessions. Enhances AI-assisted development by maintaining project-specific knowledge and conversation history.
Each quote links to its source on LinkedIn. Nothing edited, nothing fabricated.
“Lukasz was responsible for one of the most critical and technically demanding aspects of our delivery: deploying applications from unclassified development environments into high-side classified networks. This required deep technical expertise across infrastructure and security tooling, as well as a thorough understanding of the stringent compliance and assurance requirements that come with working in a Defence context. His expertise was valued even outside our team, across the wider MoD infrastructure. I would not hesitate to recommend him to any organisation looking for a senior Infrastructure or DevSecOps Engineer, particularly one operating in secure or regulated environments.”
“I've worked with Lukasz for nearly four years at the Ministry of Justice. He has excellent knowledge of cloud-platform and web-ops work, and has definitely been the go-to person whenever I needed expertise in this area. Alongside his established knowledge he is keen to try new ideas and keeps up to date with new technologies. He has been a good mentor for new members of staff. I would definitely recommend Lukasz as a much-respected member of our team.”
“I've worked alongside and been led by Lukasz since the start of my employment at the Ministry of Justice. In my professional career so far, I've yet to be led by someone who can identify an issue, establish its criticality and resolve it as quickly as Lukasz. Truly one of the best engineers and leaders in the industry.”
“Lukasz and I worked together at Hailo on a number of projects and he demonstrated a broad range of skills across operations and development. Self-directed, self-motivated, and results-driven.”
“A knowledgeable, responsible, and dependable systems engineer. On the various occasions when I've had to interrupt him with infrastructure, security, or other questions, he's always been attentive and responsive — resolving problems quickly or getting back in a timely fashion. A pleasant colleague with a great sense of humour.”
“Great experience in high-traffic environments. Solid skills, hands-on, problem-solver, with a wicked sense of humour — an extremely valuable asset for any company. I had the pleasure to work closely with him on systems hardening and cloud cost reduction, and he played a decisive role in achieving great success. Highly recommended.”
“His enthusiasm, coupled with his extensive skill-set, doesn't only get tasks 'done' — it gets them done correctly. With his ability to understand a project and his systems expertise, he delivers on time and helps reshape and advance the project as it goes. Lukasz's wealth of experience and attitude to work is rare: the desire not just to do what needs doing, but to be ready for what may need doing. I would have no hesitation in working with him again.”
“Responsible, highly reliable, and goal-oriented, with extensive technical knowledge that reaches far beyond the normal systems-engineering remit. Self-motivated, highly helpful, with a positive attitude — always happy to take on new challenges, even outside his scope of duty. The tougher the challenge, the more enthusiastic he is about it.”
Longer-form pieces on infrastructure decisions, incident post-mortems, and the boring-but-necessary parts of keeping production systems honest.
If you’ve ever wanted Claude Code or Cursor to actually read and write your knowledge base — not just chat about it — you’ll get this.Continue reading...
If you have ever worked with port forwarding in Kubernetes, you must know all the associated pains by now. Don't worry — I found an actual…Continue re...
Have you ever wondered how to monitor your GraphQL server in an easy, non-disruptive way? I think I’ve found an answer.Continue reading on ITNEXT »
Everyone should know about the six do’s and don’ts of the code review process.Continue reading on ITNEXT »
Subscribe via RSS
I don't take on more than I can keep my head around. If that sounds like a constraint worth working within, it is.
London, United Kingdom · responding to [email protected]