Martin Hlaváč

linkedin.com/in/martinhlavac · github.com/mhlavac · mhlavac.net · [email protected] · +49 176 342 75547 · Berlin area, Germany · Czech citizen (EU work rights)

Engineering leader: turns AI and platform investment into engineering and commercial outcomes.
16+ years engineering, 9+ leading teams and organisations.

I got happily addicted to agentic coding in June 2025. Junie → Gemini → Claude Code, all starting with a 15-minute run that actually delivered functional code. I gave a public talk at HVLtech, presented at Flink, and enjoyed it more and more ever since ;-)

Leadership: org design, growing EMs & Staff ICs, stakeholder partnership, AI transformation at scale, Platform & DevEx, FinOps & vendor negotiation, hands-on with AI

AI: agentic adoption (Claude Code, Codex, Copilot), MCP ecosystems, multi-agent orchestration, adoption telemetry · Platform: GCP, Kubernetes, Terraform, Datadog, FinOps · Code: Go, Python, TypeScript, PHP, Java, all behind automated quality gates

Selected outcomes

• Led Flink's agentic-AI adoption across ~80 engineers / ~13 teams: GitHub Copilot first, then Claude Code from pilot to daily default, 10+ MCP plugins plus the plugin marketplace, adoption instrumented end-to-end (OTLP into Datadog). Recognised by the CTO as the primary driver of making AI practical for developers.
• Turned around the Consumer domain, eNPS −1 → +16 across 4 teams: teams own their KPIs and roadmaps, engineers present Domain Business Reviews to leadership.
• Cut cloud and observability spend by €250k+/yr through usage optimisation and direct vendor negotiation; governs a ~€2M R&D tax-credit process (58 projects) with Finance.
• Developer platforms since Wayfair Berlin: built PHP Platform from zero for 1,000+ engineers; at Flink ran the entire Platform group and founded the DevEx team.

Experience

Flink SE (Berlin) Jun 2023 → now

Associate Director of Engineering, Consumer Jan 2025 → now

Leading the Consumer domain (checkout, pricing, ordering, post-order: up to 4 teams, ~25 engineers, 4 EMs) while driving Flink's engineering-wide AI transformation. Reporting to the CTO.

Product delivery (with the teams, working closely with business stakeholders):

• Segmented Prices & Delivery Fees (+1.9% cart-to-checkout CVR); Weight-Based Pricing.
• Next Day Delivery & Delivery Options rewrite.
• PayPal Express Checkout, Retail Media integrations, Sovendus on the delivery screen.
• Help Center flows that improved deflection and NPS; the Loyalty Badges program.
• GDPR health: data-removal remediation across 86 integration destinations under a journalist deadline.

AI transformation (engineering-wide):

• Ran Claude Code from pilot (Aug 2025) to daily default: licensing and cost governance, central enablement repo, hands-on build-along sessions; seeded power users, let usage grow.
• Led by example: demoed test-validated development with Claude, opening a full PR in 30 minutes on work estimated at days; the demo won buy-in from CTO to engineers.
• Started the Flink AI marketplace (Datadog MCP, BigQuery analysis skill, first design skill); asked teams to expose internal tools as MCP for agents: less toil, faster troubleshooting.
• Cross-platform experiment: ship once, reuse the implementation as spec for other platforms. Ports: Android in 1.5 weeks (6 on iOS), web in under a day (4 weeks on iOS).
• Prototyping beyond engineering: non-engineers prototype on the real website via ephemeral previews; the Jira skill breaks business ideas into initiatives and epics agents can act on.
• Originated and prototyped Flinkpedia, agentic search across Google Docs, Confluence, and GitHub (6,500 documents, retrieval ~4 min → ~30 s); designed Flink DNA, agent-consumable service knowledge.
• Drafted the AI Commerce strategy with Commercetools and TalonOne, presented to leadership (Feb 2026); in build-out for UCP and mainstream agentic commerce.

Domain leadership:

• Rebuilt planning and reporting with the Product Director; drove the eNPS turnaround and better OKR understanding across the org.
• Grew the EM bench: expanded three EMs into larger scopes and covered a team as interim EM while running the domain.
• Commercial ownership: vendor portfolio (TalonOne, Twilio, Cloudinary, Segment, Contentful, Commercetools; renewal countered with €238K GMV-impact data); leads work-capitalization in Engineering: untangled the 2024 audit with external auditors, automated it as a Claude skill.

Senior Engineering Manager, Platform Jun 2023 → Jan 2025

Owned the entire Platform group from day one, in a post-layoff environment: Infrastructure, Developer Experience (language tooling), SRE / Operational Excellence (8 engineers serving ~10 product teams).

• Reliability ownership: started postmortems for P1s within Platform, weekly Platform Health Checks, on-call deploy bypass, near-100% service-ownership accuracy.
• Cost program: Datadog €250k → €180k/yr (custom-metrics cull, renegotiation, weekly EM cost reports, anomaly detection catching incidents before owning teams); GCP −€40–70k/quarter (Spot VMs, underutilised-resource reporting, unused-project cleanup).
• Founded the Developer Experience team from zero: charter, hiring, promoted its EM from within; first paved roads: contract testing, Go standardisation.
• Introduced 20% self-directed Fridays: engineers shipped the Platform Portal (our IDP: common actions, service catalog, org chart), Microservices Bootstrap, Flinkit Web, 5× faster IaC pipelines, automated dependency updates.
• Ran Auxiliary Engineering: platform engineers embedded with product teams, gathering feedback and driving continuous-deployment adoption.
• Introduced a transparent quarterly planning process, later adopted by the Data teams and still in use; established the company-wide EMs meet, spawning several cross-EM initiatives.
• Led the cross-domain SWAT on a customer-impacting iOS TLS bug, coordinating internal teams plus Cloudflare and O2 to root cause and fix.
• Started the org's AI-adoption arc pre-Claude: GitHub Copilot rollout from own RFC to ~90 engineers for assisted coding, code reviews, and codebase Q&A.

Wayfair (Berlin) · Senior Software Engineering Manager Feb 2017 → Apr 2023

Senior Engineer in Transportation, then EM and Senior EM in Engineering Effectiveness (internal developer platform) for thousands of engineers: PHP Platform, later Java Platform and Self-Service Orchestration.

• Founded the PHP Platform team from scratch for 1,000+ PHP engineers; grew to 15 engineers + 1 manager across the three teams.
• Cut the ~15M-LoC monolith CI pipeline from 40+ min to 5 min; project templates cut new-service provisioning from days to minutes.
• Introduced Docker to the company: made the case to the then-CTO, ran the containerization PoC (monolith plus databases), started automated e2e tests against real databases.
• Drove the monolith → microservices-on-Kubernetes transition (libraries, images, Helm charts, templates, first decoupled PHP service) with dozens of teams; led guilds across EU and US.
• Replaced internal solutions with open source (Symfony, Doctrine DBAL), drove Composer adoption and automated static analysis; upgraded the 15M-LoC monolith through 2 major PHP versions off unsupported releases, with automation for future upgrades.
• Java Platform: Spring Boot and Actuator adoption, centralised Maven POMs, automated dependency updates; Auxiliary Engagements converting PHP to Java; experiments with Gradle, Kotlin, GraalVM; formed the Java Monthly meet for teams migrating to Java.
• Self-Service Orchestration: engineers provision infrastructure and act on production resources through tooling with governance and an audit trail.

Earlier

CTO & Co-founder, Cosmic Beacon (2015–16, proptech; team of 4, product from zero) · Full-Stack Developer, Customer Alliance (2013–15) · Developer → Team Lead, Publero (2011–13).

Languages & education