Machine Learning Engineer

About ve2ventures

 We're ve2ventures, a venture studio that invests in external startups and also builds our own

 companies from the ground up. We're a mix of entrepreneurs, product people and tech experts

 who love building innovative products. Together, we're building the world's largest sports

 and media tech ecosystem.

 In 2 years, we've grown to a team of 60+, working on over 20 ventures and supported by 35

 investors. As our startups grow, so does our need for strong technical talent. That's where

 you, our new Machine Learning Engineer, come in. You'll join the data science team behind

 PadelPlay, our padel performance platform. It turns data from a wearable sensor into shot

 detection, technique feedback and AI coaching for thousands of players.

The Role

 As a Machine Learning Engineer, you'll own models from the first look at the data through to

 production. You'll build, evaluate and ship the ML and AI features players use after every

 session. You'll work closely with the app, backend and product teams, and as you grow you'll

 take on more of the technical decisions about how we model, deploy and monitor what we build.

What You'll Do

 - Own the full ML lifecycle: data collection and labelling, feature engineering, training,

  rigorous evaluation, deployment and monitoring.

 - Design evaluation that holds up in the real world. That means testing on users the model

  hasn't seen, and being honest about what the numbers can and

  can't tell us.

 - Build and maintain the data pipelines and serving infrastructure behind our models

  (serverless functions, queues, object storage, SQL/NoSQL databases) across dev, staging and

  production.

 - Build LLM-powered features such as our AI coach: prompt design, tool calling, retrieval,

  multi-language output and guardrails, using APIs from OpenAI, Anthropic and TypesafeAI.

 - Use AI-assisted, agentic development in your own workflow (e.g. Claude Code) to move fast

  without cutting corners.

 - Turn product questions from the PM and coaches into clear experiments, and explain the

  results to non-technical stakeholders.

 - Help improve our tooling, reproducibility and code quality as the team grows.

What You Bring

 Required

 - 3–5 years of experience as an ML Engineer, Data Scientist or similar, with models you've

  taken to production.

 - Strong Python and the standard ML stack (pandas, NumPy, scikit-learn; PyTorch or similar).

 - Solid grounding in classical ML and evaluation: cross-validation, avoiding leakage,

  handling class imbalance, choosing the right metric.

 - Experience deploying and maintaining models or data services on a cloud platform (Azure preferred ).

 - Hands-on experience with LLM APIs (OpenAI, Anthropic or similar).

 - Comfort with SQL and working with messy, real-world data.

 - A strong sense of ownership, a pragmatic mindset, and clear written and spoken English.

Nice to Have

 - Experience with time-series, sensor or wearable data (accelerometer/gyroscope, signal

  processing).

 - Sports, health or fitness tech experience.

 - Model explainability (e.g. SHAP) and turning model output into feedback people can act on.

 - LLM agents, RAG or LLM observability tooling.

 - Serverless and container-based deployment, CI/CD, and infrastructure as code.

 - Experience in a startup or small, fast-moving team.

Why This Role Is Interesting

 You'll work on a product where ML is the product. Your models run on real players' sessions

 every day, and the results go straight into what they see in the app. The problems are hard

 in the good way: noisy sensor data, a small number of expert reference players, and feedback

 that has to be accurate and useful to an amateur on court. You'll have real ownership, a

 short path from idea to production, and an AI-native workflow you can help push further.