woehrl.net

Head of Engineering, Cloud Platform & AI · PlentyONE GmbH

Lukas Wöhrl

I lead Cloud Platform & AI at PlentyONE, five teams and 30 people at peak, and own core product: the marketplace integration platform and the public Partner API. Before that I reported to the CEO as Director of Operations and co-designed the product and engineering organisation with the founder. 20+ years in software, from healthcare systems to e-commerce SaaS, including Microsoft.

Kassel, Germany · Remote · German & English

Lukas Wöhrl
200+person product & engineering org, designed with the founder
~100company-wide production releases per working day
40%↓monthly cloud compute cost since 2023, while usage grew
3×faster support response & resolution

What I work on

Building engineering organisations that ship the right product, reliably: clear structures, strong leads, shared knowledge and an architecture that keeps teams independent.

ProductEngineeringOrganisation
Where product, engineering and organisation meet

Teams, leaders & organisational design

Lead five teams, 30 people at peak, from AI to Infrastructure. Hired 20+ engineers, grew five into leaders and, with the founder, designed the 200+ person product and engineering organisation, including its first areas, teams and managers in 2022.

  • Leadership
  • Mentoring
  • Organisational design
  • Multicultural teams

Product & customers

Delivered a marketplace integration platform and a public Partner API in four months with a team of eight: consistent logic and UI across marketplaces, one front door for partners. Measure what users do and ship behind feature flags, so we learn fast and roll back early.

  • Product thinking
  • E-commerce
  • Marketplaces
  • Partner API

Strategy, cost & vendors

Own the company-wide cloud budget, the people budget for my area and the architectural direction. Cut monthly compute cost by 40% since 2023 while usage grew, keep AI spend in check and lead contract negotiations with AWS and other vendors.

  • Strategy
  • Budget ownership
  • FinOps
  • Due diligence

Developer experience

Made the right way the easy way: one internal CLI, ready-made dev environments that cut setup time by 90%, single sign-on everywhere and permissions requested through a reviewed pull request, not a ticket.

  • Developer tooling
  • SSO
  • Self-service
  • GitOps

Cloud platform

Moved the platform behind 10,000+ customer systems from EC2 to Kubernetes, across multiple clusters, availability zones and regions, with GitOps and an observability stack we own end to end. Releases roll out in rings that stop automatically on error spikes; my teams also run on-call, incident response and disaster recovery.

  • Kubernetes
  • Argo CD
  • AWS
  • Reliability & on-call

AI

Lead the AI team and the company's AI strategy: deterministic first, AI only where it earns its place on cost, speed and debuggability. Rolled out AI coding agents company-wide, used daily by 93% of engineers over three months, plus a support chatbot on AWS Bedrock and living architecture docs generated from recorded engineering sessions.

  • MCP
  • AWS Bedrock
  • LLMs
  • Knowledge sharing

Experience

  1. Dec 2020 – present

    PlentyONE GmbH · formerly plentysystems AG

    Head of Engineering, Cloud Platform & AI Jul 2022 – present

    Five teams, 30 people at peak, in several countries: tech lead managers, SREs, engineers and data analysts · reporting to the CTO

    • Shipped the marketplace integration platform and the public Partner API in four months with a team of eight: consistent logic and UI across marketplaces for customers, one front door for partners.
    • Represented engineering in the technical due diligence for the investment from private equity firm PSG, taking the investors through architecture, security, cost and delivery.
    • Own the company-wide cloud budget and my area's people budget; cut monthly compute cost by 40% since 2023 while usage grew and lead vendor negotiations.
    • Moved the company to cloud-native infrastructure with ring-based progressive delivery and feature flags; product teams self-serve compliant AWS resources via Kubernetes and Terraform.
    • Own security and compliance with my teams: SOC 2, TÜV and Amazon audits, single sign-on and access management, penetration testing, vulnerability management and security training.
    • Hired 20+ engineers and grew five of them into leadership roles.
    • Led a restructuring that reduced headcount: decided with the leadership team where to cut and kept ownership and delivery clear.

    Director of Operations Dec 2020 – Oct 2022

    Organisational development across product, engineering and support · reporting to the CEO, working closely with HR and the executive team. From Jul 2022 alongside Head of Engineering, handing areas over to new leaders

    • Co-designed the 200+ person product and engineering organisation with the founder, including the 2022 reorganisation that gave a flat organisation its first management structure.
    • Designed, rolled out and measured a new support process that made response and resolution 3× faster; analysed support load by product area to show where the product hurt customers most.
    • Introduced user activity tracking, so product decisions were based on real usage.
    • Built developer tooling that cut setup time by 90%.
    • Built the career framework and performance management with HR and the executive team.
    • Moved product and engineering to DevOps processes; introduced a company-wide knowledge base and a new onboarding experience.
  2. Mar 2019 – Dec 2020

    Microsoft

    Premier Field Engineer for Developers Mar 2019 – Dec 2020

    Hands-on technical consultant embedded with customer engineering teams

    • New customers and new problems every week, often on site with one day to understand the problem and find a solution. Rated above expectations and promoted within my first 12 months, in a team of 50.
    • Advised development teams on C#/.NET Core, DevOps/ALM, Kubernetes and SQL Server, on-premises and in Azure.
    • Led cloud migrations from on-premises and AWS to Azure.
    • Delivered technical workshops and training for customer engineering teams.
    • Received the Microsoft Services #BetterTogether Award (2020), given to teams whose work turned business priorities into customer outcomes.
  3. Mar 2012 – Feb 2019

    RECOM GmbH · now Thieme RECOM

    Lead Software Engineer Mar 2012 – Feb 2019

    Software Engineer → Lead Software Engineer, team lead from 2016

    • Built digital nursing documentation for hospitals and care facilities in German-speaking countries, based on the ENP nursing terminology and integrated with hospital information systems.
    • Led a team of software engineers from 2016, owning architecture, mentoring and internal frameworks.
    • Moved the organisation to agile practices with continuous integration and deployment.
    • Modernised 25-year-old legacy code; built REST, SOAP and HL7 interfaces to hospital systems; tuned .NET and SQL Server performance.
  4. 2005 – 2011

    Early career

    Internships, apprenticeship & military service 2005 – 2011

    • Intern at Microsoft (Cloud Computing, Aachen), military service in the navy (commendation for outstanding performance), IT apprenticeship and system administration at a bank.

Education

T-shaped

Broad and deep

Broad enough to work with every part of a company, deep enough to make and challenge the hard technical calls.

Breadth: where I've worked hands-on

  • Software engineering
  • Product management
  • E-commerce
  • Healthcare IT
  • Data engineering & BI
  • Observability
  • Incident management & SLOs
  • Security, identity & compliance
  • AI & LLMs
  • Customer support
  • Consulting
  • Organisational design
  • HR & people processes

Depth: where I go all the way down

  1. Organisational design & engineering leadershipTook a flat organisation to areas, teams and managers; career framework and performance management with HR; M.A. in organisational development.
  2. Kubernetes & platform engineeringFrom GitOps and autoscaling to node-level tuning; contributor to Argo CD, Karpenter, metacontroller and OpenTelemetry.
  3. Developer experience & self-serviceSSO everywhere, permissions via pull request, one CLI for daily work.
  4. AWS architecture & costHands-on across the AWS ecosystem, from EKS, VPC and IAM to data, messaging and serverless; Azure from my Microsoft years.
  5. Go, .NET, C++ & TypeScriptGo for Kubernetes operators, C# and .NET across a decade of product work, C++ as the main external contributor to Yoga (the layout engine behind React Native), TypeScript for infrastructure as code; profiling and tuning at every layer.

How I build

Complexity has to earn its place.

I've built systems more complex than they needed to be, and learned the most from running them. So I start boring: a service and a database, with clear boundaries that leave room for where the product is heading. Scaling, caching and splitting into services come later, when measurements show tuning has run out. For most products, microservices on day one are premature optimisation. At PlentyONE, the core product stays one system with clear module boundaries; we split out a service only when scaling or ownership demands it.

As a leader I also see the other side of the bill: every extra component adds cognitive load, onboarding time and on-call burden. I've been on call myself, so I want systems a team can understand, own and fix quickly, with a short path from “something broke” to “I know why”. Simple architecture lets teams move on their own.

  1. Service + database start here
  2. Profile & tune
  3. Scale up, then out
  4. Add a cache
  5. Split into services
Add the next layer only when the one below runs out

How I lead

Teams that don't need me in every decision.

The same rule holds for organisations. More teams and more people aren't a success in themselves; every extra team adds handoffs, coordination and work in progress. I'd rather have fewer, stronger teams with clear ownership, and I make the hard calls that come with it, openly and fairly.

I hire for judgement and ownership, not titles. My karate sensei used to say: “The belt only holds the jacket together.” It took me years to understand: a title doesn't make you good at something, practice does. So teams own their systems end to end, from roadmap to on-call, and decisions sit where the knowledge is. Nearly all technical decisions are made by the teams themselves; they come to me for strategic and architectural direction.

With the executive team, product and HR I work as a partner: shared goals, honest trade-offs and forecasts from delivery data. Together we built the career framework and performance management, which I still help run.

My teams, and the leaders I work with, span several European countries and time zones. That taught me to make expectations explicit, write things down and build trust on purpose. I've led remotely for years and learned how to keep people connected without sharing an office, and when it's worth bringing everyone together in person.

The leadership loop
20+ hires 5 grown into leaders Hire for judgement & ownership Grow engineers into leads Mentor 4,300+ code reviews Share 400+ Show & Tell talks Forecast from delivery data 20+ hires 5 grown into leaders 12345
  1. Hire for judgement & ownership
  2. Grow engineers into leads
  3. Mentor 4,300+ code reviews
  4. Share 400+ Show & Tell talks
  5. Forecast from delivery data

Open source

I'd rather build on the ecosystem and contribute back than reinvent it in-house: when OpenTelemetry or metacontroller lacked something we needed, we fixed it upstream instead of building and maintaining our own tooling.

424 commits merged into 33 open-source projects since 2014

Contributed to

Created

  • pod-pacemaker Go

    Kubernetes operator that controls the rate at which pods start on a node.

  • aws-auth-jwt Go

    Authenticate via an IAM role and get a signed JWT for authorisation.

Talks

  • Observability first: plentysystems' cloud native journey

    AWS Summit Berlin, May 2024, with Klaus Kurz (New Relic): modernising the infrastructure behind a cloud-native commerce platform, putting data in the hands of every decision-maker and making the cultural change stick.

  • #NoEstimates

    Why estimating less can help teams deliver more predictably. Slides (PDF)

  • Azure Pipelines from zero

    Live demo of a basic CI pipeline at Webmontag Kassel. Demo code

Let's talk

Happy to compare notes with founders, CEOs and boards of B2B SaaS companies on engineering leadership, organisational design, product, platform engineering, AI and cloud cost. If you're building something where complexity has to earn its place, get in touch.

Imprint & privacy

Lukas Wöhrl

Franz-Vetter-Straße 4b
34121 Kassel, Deutschland
E-Mail: lukas@woehrl.net

Umsatzsteuer-Identifikationsnummer: DE 261 806 524
Inhaltlich Verantwortlicher gemäß § 18 Abs. 2 MStV: Lukas Wöhrl