Role Overview
Join a VC-backed HealthTech startup to build and productionize ML models that power predictive health insights from radar sensor data. You will work on reliable, real-world ML systems used in production, collaborating closely with hardware, backend, data engineering, and product teams.
Responsibilities
- Build ML models using XGBoost, ensemble methods, anomaly detection, and time-series techniques.
- Develop feature engineering pipelines for radar signals, point-cloud, and sensor data.
- Work on movement, pose, and activity estimation using radar-based data.
- Build training, evaluation, and inference pipelines using Databricks.
- Perform exploratory data analysis to improve model performance and identify edge cases.
- Evaluate models using production metrics such as precision, recall, false alarms, and detection latency.
- Monitor and improve production ML systems across devices and deployments.
- Write clean, production-grade Python code and collaborate with hardware and data engineering teams.
Requirements
- 3–4 years of experience building and deploying ML systems in production.
- Strong Python programming skills.
- Solid understanding of feature engineering, model training, cross-validation, and model evaluation.
- Experience with XGBoost, Random Forests, Gradient Boosting, anomaly detection, or time-series models.
- Good SQL skills with experience using PySpark, Pandas, or Databricks.
- Experience working with sensor, IoT, time-series, point-cloud, or computer vision data.
- Familiarity with Spark, Delta Lake, or modern data engineering workflows.
- Strong debugging, analytical thinking, and ownership mindset.
Nice to Have
- Experience in HealthTech, IoT, or safety-critical systems.
- Exposure to Computer Vision, pose estimation, or object tracking.
- Experience with MLflow, model monitoring, or experiment tracking.
- Knowledge of ONNX, edge deployment, or latency optimization.
- Experience with Kafka, Spark Structured Streaming, or real-time inference systems.
Skills
- Python
- Databricks
- XGBoost
- PySpark
- SQL