Cristian Avramescu

Senior Engineering Manager

Almere, Netherlands (Amsterdam area) | +31 6 14 83 58 74 | cristian.avramescu@outlook.com
linkedin.com/in/cristian-avramescu-0ab3792b | EU citizen, no sponsorship required

Professional Summary

I step into the operational chaos of scale-ups and build organisations that pull together and perform. I work the way a sports coach takes an underdog roster to a championship: I pick for attitude, values and coachability over raw technical skill. I give driven people the room to learn, take ownership and deliver real outcomes, all pointed at the company's North Star. A strong moral compass and a relentless work ethic keep my decisions principled, and curiosity keeps me adaptable. I connect the technology side of a business with the human side, and that's how I turn messy, scattered resources into autonomous, championship-level teams.

• Deep technical judgement. Seven years as an engineer before moving into management. Founded an analytics platform that now handles ~900M events a day and set its architecture as it scaled (job cycles 8–12 h → 1–2 h); moving its storage from Redis to S3 halved its running cost.
• People who grow. Developed a QA into a team lead running her team in less than 6 months, an under-performer into a senior engineer in 12, and data engineers with no software background into platform engineers.
• Programmes across teams. Led the main search platform's move from EC2 to Kubernetes — 15 engineers across 3 teams, mostly people who didn't report to me — and took two teams to fully agentic development in two months (~20% higher throughput).

Core Competencies

Leadership: Engineering management, cross-team programme leadership, organisational design, hiring and onboarding, performance management, coaching and mentoring, succession planning, career development and promotions, stakeholder and escalation management

Architecture and platforms: Software architecture, distributed systems, data platforms, cloud architecture and cost optimisation, platform modernisation and migration

Engineering effectiveness: CI/CD and release automation, test automation, observability, GitOps, agentic / AI-assisted development

Technical foundations: Java, Spring Boot, Scala, Python, SQL, Apache Spark, Trino, Kubernetes, Helm, ArgoCD, Argo Workflows, AWS, GCP, Prometheus, Grafana, Claude Code

Professional Experience

Software Development Manager — Crownpeak (acquired by Rezolve AI, 2025), Amsterdam

Apr 2023 – Present

E-commerce search and merchandising platform: ~900M events/day, 25 microservices, 110 cloud accounts. Across acquisitions the company grew from ~150 to ~1,400 people while engineering shrank from ~50 to ~35; over the same period my scope grew from one team to two teams plus a three-team programme. Report to the Head of Engineering.

• Led the platform-wide replatform from EC2 to Kubernetes (late 2024 – 2026), targeting a ~30% reduction in platform running costs. Grew it from a 3-engineer project covering a third of the platform into a programme of three development teams — 13 engineers plus 2 infrastructure engineers, most not reporting to me — spanning ETL, Search & Navigation, orchestration and Suggest. Set the technical direction, drove its expansion to the whole platform, and owned sequencing and technical arbitration across the teams through the bulk of the build.
• Led two teams through the move to agentic development in two months (2026). From ad-hoc Copilot use to spec-driven development with Claude: set the requirement for an engineered harness and agreed its guardrails, which the teams built (double-loop TDD enforced on agents, AI-then-human PR review, human checkpoints on spec and plan). ~20% throughput increase on initial measurement after two months.
• Replaced the Pentaho ETL platform (ETL 2.0, 2024–2025). Led architecture and technology selection (Apache Hop, Argo Workflows/Events, ArgoCD) while training two data engineers with no software background to build production software. 40–80% faster and 10–50% cheaper across like-for-like comparisons (same catalogue, customer and configuration), varying by workload.
• Unblocked a stalled ownership handover with the Heads of Engineering and Technical Consultants, moving my data engineering team from 20% project / 80% maintenance to 75% / 25% in eight months — the change that made the replatform possible.
• Managed 11 people across two teams at peak (2024), including a team lead with her own reports; today 8 engineers. Trained a senior QA into my successor as Insights team lead, running the team largely unaided in under 6 months; mentored a second team lead outside my reporting line.
• Rebuilt the reporting API on Trino (2023–2024): report loads from 5 s – 2 min to 1.5–10 s, flat regardless of date range; data-correctness investigations cut from ~3 days to half a day.

Software Development Team Lead — Attraqt (acquired by Crownpeak, 2022), Amsterdam

May 2020 – Apr 2023

Led the Insights analytics team: inherited 4 developers + 1 QA, grew to 10 direct reports (developers and QAs) by 2022, plus a second team and product absorbed in 2021. The product had to run as close to zero cost as possible.

• Halved the platform's running cost (−50–55%) by moving the event store from Redis to S3, removing a 4-hour retention limit at the same time.
• Hired two juniors end to end: posted the roles, screened 100+ CVs, designed the technical assignment and ran the full loop without recruiter support. Both promoted to regular within a year.
• Grew capability from within: turned around an under-performing developer through a direct conversation instead of a PIP (promoted to senior 12 months later), and cross-trained an absorbed team of 3 .NET developers onto Java, Scala, Spark and Kubernetes.
• Designed out single points of failure: capped work at 2 concurrent projects, ran a full pair-programming rotation, introduced epic owners and a production-support rotation. When engineers later left, the team lost capacity, not capability.
• Custom job scheduler (2022): job runtimes from 8–12 h to 1–2 h regardless of volume; backfills from 5 days × 2 engineers to ~1 hour × 1 engineer. Rebuilt the Spark jobs: +40% throughput on equal or fewer resources; new-metric lead time from 1–2 months to 3–5 days.
• Release cadence from every 2–3 months to every sprint, with the company's first fully automated end-to-end pipeline — and the first development team in the company to own its own monitoring.

Software Engineer / Senior Software Engineer — Fredhopper (acquired by Attraqt, 2017), Amsterdam

Nov 2016 – May 2020

• Founded the company's first analytics platform (2018) in its 18-year history: prototype live in 3 months, then redesigned ingestion so customer integration became trivial.
• Ran the team almost alone for several months in 2019, then helped hire and onboard 3 engineers. Built a GCP activity pipeline (Dataflow, BigQuery) with an acquired company's squad.
• Owned global merchandising rules end to end, collapsing customer setups from thousands of rules to dozens; cut list interactions from up to 30 s to sub-second.

Research and Development Engineer — Esker SA, Lyon, France

Sep 2013 – Sep 2016

• Built features for a sales-order-processing web application (C++, C#/.NET, JavaScript); owned automated tests (Selenium, QUnit) and the Jenkins CI system.

Independent Technical Work

AI-operated homelab (personal, 2024 – Present): run a 7-node, ~70-service Kubernetes platform (GitOps with ArgoCD and Helm, Prometheus/Grafana/Loki) mainly through AI agents, under an autonomy policy and hook-enforced guardrails I designed; built and later retired a 15-agent fleet when its cost outweighed its value.

Education

Master's degree, Computer and Information Science — INSA Lyon (Institut National des Sciences Appliquées), France, 2008 – 2013

Languages

Romanian (native), English (bilingual), French (bilingual), Spanish (elementary), Dutch (beginner)