Curated Overview

My CV

Engineering Manager Snapshot

A fast-read view of the same story as the PDF: applied ML systems leadership, research-to-production delivery, model evaluation, reliability, and practical AI standards.

🧠 Positioning

ML Engineering Manager for applied ML systems and real-time data products

I lead cross-functional teams that translate research-grade models, noisy data, and ambiguous product requirements into reliable production systems. At Vortexa, that means managing 6 direct reports, leading a 10-person team, and keeping model quality, evaluation, reliability, and stakeholder alignment close to the engineering work.

  • 👥 Team leadership
  • 🧠 Applied ML systems
  • 📐 Model eval/replay
  • ⚙️ Model serving
  • 🛡️ Client-facing reliability
  • 🏛️ AI standards
Download PDF CV Open Experience Timeline
Christos Hadjinikolis presenting at Big Data London in 2018

🧭 What I Actually Do

  • Build and manage technical teams Direct management, hiring, mentoring, reviews, progression, onboarding, delivery accountability, and cross-functional operating rhythm.
  • Move ML work from research towards production PyTorch sequence/transformer models, model-serving workflows, MLflow, model/data versioning, evaluation gates, replay, and monitoring.
  • Turn ML ambiguity into operating discipline Batch/online evaluation loops, failure-mode analysis, domain-expert feedback, product semantics, and prediction trust.
  • Make applied AI systems inspectable Tool boundaries, schema validation, approval gates, durable state, traces, and runtime evidence for humans and LLM-assisted workflows.

📌 Evidence Behind The CV

  • People: manage 6 direct reports across Product, SME analysis, Data Science, and Data Engineering; lead a 10-person cross-functional team accountable for model quality, stakeholder alignment, reliability, and delivery maturity.
  • Estate: own engineering strategy and delivery for a live ML/data estate turning roughly 6M vessel-position records/hour into production intelligence for 13.5K monitored vessels.
  • ML delivery: led 0-to-1 research-to-production delivery for destination and arrival-time sequence/transformer models in PyTorch.
  • Evaluation: established batch/online model-evaluation and replay loops, analysed failure modes with domain experts and Product, and converted findings into model, data, and interface improvements.
  • Reliability: protect production trust through Kafka Streams-to-Flink migration, monitoring, fallback/rollback paths, shared on-call, and MTTR kept under 30 minutes.

🧪 Applied AI & Tooling

  • Promet: private applied GenAI project shaping hands-on work around voice, memory, tool use, streaming interaction, local runtimes, Hugging Face-backed speech assets, schema validation, approval gates, durable state, traces, and replay/evaluation.
  • skeleton-replay: public Python tooling that turns script/pytest runs into traces, architecture snapshots, workflow evidence, and replayable reports for review, debugging, onboarding, and LLM-assisted code understanding.
  • Skeleton Replay plugin: PyCharm/IntelliJ workflow that brings runtime evidence and source navigation into the IDE.
  • dynamicio: published PyPI library for making I/O seams and local/dev/prod dataset switching explicit in ML/data workflows.

📚 Career Snapshot

  • 12/2020–present · Vortexa, London
    Engineering Manager / ML Systems Lead owning engineering strategy and delivery for a live ML/data estate, managing 6 direct reports, and leading a 10-person cross-functional team around model quality, reliability, stakeholder alignment, and delivery maturity.
  • 04/2016–12/2020 · Data Reply, London
    Senior Consultant and first London spin-off consultant; grew from data scientist into ML engineer while supporting team growth, client delivery, mentoring, and project leadership across Vodafone, CNHi, and UBS.
  • 2010–2016 · KCL, UCL, GSM, David Game College
    Teaching and academic roles across computing, AI, software, and data subjects.

🏛️ Standards & Research

  • Since 10/2024 · UCL
    Associate Researcher helping students connect AI standards, the AI Act, auditability, explainability, and practical AI adoption.
  • Since 01/2021 · ISO/CEN-CENELEC JTC 21 WG3
    Committee Expert Member contributing to AI standards aligned with EU policy and international norms, with emphasis on auditability, model/data versioning, explainability, and safer adoption.

📐 Core Principles

  • Production is the only truth.
  • Models need evaluation, replay, monitoring, and graceful failure paths.
  • Responsible AI is partly an engineering discipline: evidence, auditability, ownership, and human accountability.
  • System quality should come through clear ownership, measurable interfaces, and repeatable practice.

🎙️ Talks & Interviews

  • 2023 Agile in Action podcast interview on the Vortexa journey and agile data science.
  • 2022 ODSC talk on dynamicio, a published PyPI library for abstracting I/O in ML systems.
  • 2020 iunera interview blog on the agile approach in data science.
  • 2020 Big Data Warsaw talk on monitoring communication and trade events as graphs.
  • 2018 Connected Data London panel and Minds Mastering Machines talk.

🎓 Education & Credentials

  • Ph.D. in Computer Science · King’s College London
    Persuasion dialogues, opponent modelling, knowledge graphs, Bayesian techniques, and formal semantics.
  • Diploma (BEng) in Computer Engineering · University of Thessaly
    Polytechnic training with a strong focus on mathematics and artificial intelligence.
  • Selected certifications
    AWS ML Specialty, Google Data Engineer, Process Mining, Graph Analytics for Big Data, Neo4j, and Elasticsearch.