Portrait of Jedrzej Nowak
Systems architect · engineering leader · builder

Poland · global remote · 2026

Calm systems.
Clear ownership.

I’m Jedrzej Nowak, usually called NJ. I’ve spent most of my career in places where failure is expensive: telecom, cloud platforms, blockchain, and now AI. I help companies turn technical ambition into systems and teams that keep working when things get real.

If you have a difficult system or an engineering organisation under strain, send me a note, book an introduction, or download the CV.

Current work
Building an AI product and shaping the engineering system around it.
Most useful at
The point where architecture, operations, product, and organisation meet.
Optimises for
Reliability, recovery, clear ownership, strong teams, and low drama.

How I work when the problem is ambiguous and the stakes are real.

I’m not interested in looking important from a distance. I’m useful when a company needs architectural clarity, a steadier engineering system, and someone who can stay close to the problem without becoming the bottleneck.

Close enough to help. Far enough not to suffocate.

I join design reviews, incident work, and difficult delivery conversations when it improves the outcome. Then I step back so the team keeps ownership.

Product, platform, and people are one system.

I look at organisational friction and technical friction together. Slow teams often have structural problems, not motivation problems.

Geeky, but not doctrinaire.

I care about solid design, but not as theatre. If the right answer is boring, pragmatic, and maintainable, I will take it every time.

Selected work: reducing operational drag without flattening ambition.

The pattern is usually the same: too much ambiguity, too much operational burden, or a team carrying more cognitive load than it should. My best work is reducing that load without making the system brittle.

Platform

Reduced infrastructure drag while keeping reliability expectations intact.

Reworked platform footprint, matched workloads to better-fit environments, and brought operating costs under tighter control without trading away operational confidence.

  • Bare metal and cloud rationalisation
  • Cost discipline tied to service quality
  • Less waste and fewer surprises

Organisation

Gave distributed teams a clearer operating model.

Helped remote engineers work with better ownership, healthier handoffs, and less needless coordination overhead.

Production

Stayed close to systems with real operational burden.

Worked where architecture decisions had to hold up under continuous use and scrutiny, not just in a design document.

Operations

Improved reliability in environments that punish sloppiness.

Supported telecom, infrastructure, and blockchain work where incident quality and recovery mattered more than polished narratives.

Defaults I keep reaching for when systems get messy.

  1. Design for recovery, not imagined perfection. Failure will happen. Good systems make recovery legible through clear ownership, boring runbooks, realistic SLOs, and architecture that does not require superheroes.

  2. Push authority toward the team doing the work. Teams move faster when decision rights are explicit. I provide context, pressure-test options, and clear obstacles. I do not centralise every choice.

  3. Measure the operational truth. Slide-friendly metrics are not enough. I care about cycle time, incident quality, deployment confidence, retention, and whether the architecture is getting easier to change.

  4. Choose tools for fitness, not identity. The goal is not to signal taste. The goal is to build an engineering system that a tired, competent maintainer can understand and extend six months later.

Recent chapters and the problems I was trusted with.

  1. 2025 → now

    Founding Engineer · Stealth AI Startup

    Building product from scratch, improving technical direction, and helping set the tone for how engineering should work before habits calcify.

  2. 2022 → 2025

    Head of Engineering · Spacemesh

    Led engineering through launch and steady-state operations of a novel blockchain network, while improving team shape and tightening operational discipline.

  3. 2021 → 2022

    Senior Engineering Manager & Site Manager · CodiLime

    Managed multiple cloud-native projects, mentored senior engineers, and kept delivery moving across parallel workstreams without losing technical coherence.

  4. 2019 → 2021

    Head of Engineering · Travelping GmbH

    Ran cloud-based mobile network core systems with strict uptime expectations and real-world operational consequences when things went wrong.

  5. 2015 → 2019

    Engineering Manager / Tech Lead · Mirantis

    Built large-scale OpenStack and Kubernetes automation, led distributed teams, and helped customers move from fragile installs to repeatable platform operations.

  6. 2011 → 2015

    Co-Founder / Technical Lead · Codernity

    Co-built products, a company, and the broad technical instincts that only show up when architecture, delivery, and business constraints arrive at once.

  7. 2004 → 2011

    Early career · Engineering and consulting

    Built the base layer: backend engineering, consulting, open-source work, and the systems instincts that shaped everything that came after.

If you want to talk through a difficult system, reach out.

I’m always up for a serious conversation about architecture, reliability, delivery friction, or the human side of making complex technical work actually work.