Senior Data Scientist

Senior Data Scientist

About FleetHD

FleetHD is the intelligence layer for fleet maintenance. We turn raw diagnostic, inspection, and operational data into the alerts, priorities, and insights that commercial fleet teams rely on, from the maintenance office to the shop floor to the road.

Our platform ingests and cross-references data from telematics providers, sensors, inspection reports, and other fleet systems, then processes it through a configurable intelligence engine that separates the signals that demand action from the noise. From there, FleetHD is the software fleet teams work in every day — turning that intelligence into prioritized alerts, vehicle health scoring, diagnostic insights, and resolution tracking, delivered through real-time notifications, dashboards, and mobile tools.

The Role

We’re hiring a Senior Data Scientist to build the machine learning that makes FleetHD smarter. You’ll work with rich, real-world data — vehicle diagnostics, telematics, sensor readings, and maintenance history — to build models that anticipate failures, score the confidence and severity of issues, and surface what matters before it becomes a breakdown on the side of the road.

This is a hands-on, senior individual-contributor role for someone who builds models, not just dashboards. You’ll own problems end to end — framing the question, exploring the data, training and validating the model, putting it into production, and keeping it healthy once it’s there. This is not a role where you hand a prototype to someone else to ship. If that reads as a constraint, this isn’t the right role. If it reads as the chance to shape how a company applies machine learning, it is.

You’ll report to our Senior Director of Engineering and work closely with Engineering and Product, with direct exposure to real customer problems and to FleetHD’s leadership.

What You’ll Do

Build models that ship

• Design, train, and validate machine learning models on fleet diagnostic, telematics, sensor, and maintenance data.

• Take models from exploration to production and own them there. Accuracy, monitoring, and retraining stay with you.

• Build models that anticipate failures, score issue confidence and severity, and turn high-volume signal into clear, trustworthy output for customers.

Work the data deeply

• Explore, clean, and engineer features from large, messy, real-world operational datasets.

• Apply time-series, anomaly detection, and predictive modeling techniques suited to sensor and diagnostic data.

• Take on the hard version of the labeling problem: ground truth arrives weeks late, through repair records of uneven quality. Deciding what counts as a correct prediction is part of the work.

• Build the data foundation the modeling depends on — normalizing diagnostic data across telematics providers, joining it to repair outcomes, and preserving the point-in-time history that makes an honest model possible. We have multi-year fault code history and real work-order data to start from. Turning that into a training set is early, substantial work.

Own model quality in production

• Define what you’re optimizing for and defend the tradeoff. In our world a false alarm costs a fleet a wasted inspection; a miss costs them a truck on the shoulder.

• Instrument models for drift, degradation, and calibration, and set the alerts that tell you when to retrain — and what to do when they fire.

• Explain why a model made the call it made, in terms a maintenance manager will accept. Our customers act on our output or they don’t, and an alert nobody can explain doesn’t get worked.

Shape how FleetHD does ML

• Partner with Engineering and Product to turn modeling capabilities into real product features.

• Help shape the data, tooling, and practices behind how FleetHD applies machine learning as the team grows.

• Bring a pragmatic point of view on where machine learning adds genuine value (and where it doesn’t).

What We’re Looking For

• 5+ years building and shipping machine learning models that ran in production — and that you personally supported once they were live.

• Strong Python and SQL. Day to day that means pandas, NumPy, scikit-learn, and gradient-boosted trees (XGBoost, LightGBM, or similar), with PyTorch or TensorFlow where a problem genuinely warrants it.

• Depth in predictive modeling, and ideally time-series or anomaly detection — the kinds of problems that come with sensor and diagnostic data.

• Comfort shipping models into a containerized cloud environment. We run on Azure, though experience on any major cloud works. Once a model is live, you keep it healthy.

• A bias toward the simplest approach that solves the problem. Complexity has to earn its place here.

• Fluency with model explainability techniques (SHAP or equivalent), and the ability to translate what a model is doing into plain language for people with no data science background.

• Experience with large, messy, real-world datasets, and the judgment to turn them into reliable features and signals.

• Comfort using AI-assisted development tools in your own workflow, and judgment about where LLMs genuinely help in a product like ours — unstructured repair and work-order text, for instance — versus where a traditional model is the right answer.

• The ability to own ambiguous problems end to end and communicate results clearly to technical and non-technical partners.

Background

We value strong academic training in a quantitative field — computer science, statistics, engineering, applied mathematics, or similar — and an advanced degree is a real asset in this role. It is not a requirement. We are just as interested in candidates whose depth came from years of demanding production work.

What sets a candidate apart

• Experience with sensor, IoT, telematics, automotive, or industrial and operational data.

• Background in failure prediction, reliability, or equipment health in a physical-world domain.

• Experience building data and ML infrastructure from an early stage.

• Hands-on experience with infrastructure as code and container orchestration (Terraform, Kubernetes, or similar). On a small team the line between data science and platform work moves, and it helps to be able to cross it.

• Experience in a setting where a false positive and a false negative carried genuinely different costs, and you had to make that call.

Why This Role Matters

Machine learning is central to FleetHD’s future. You’ll work on hard, real-world problems that directly affect whether trucks stay on the road. The models you build will reach some of the largest commercial fleets in North America, shaping what their teams see every day. It’s a high-impact, foundational role with significant ownership and real room to grow as the company scales.

Compensation & Benefits

The base salary range for this role is \$180,000–\$210,000. Final compensation will depend on experience, skills, and location.

FleetHD offers a competitive benefits package, including health, dental, and vision coverage, a retirement plan, and paid time off.

Equal Opportunity

FleetHD is an equal opportunity employer. We are committed to building a diverse team and do not discriminate on the basis of race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability, veteran status, genetic information, or any other characteristic protected by law. We are happy to provide reasonable accommodations for candidates with disabilities throughout the hiring process.