Applied Scientist TimeSeries

Who We Are
Striv is a startup building AI-powered wearable technology for real-time athletic feedback, with team based in Boston. We operate at the intersection of smart wearables, biomechanics, and AI, helping users detect risk earlier, train smarter, and improve performance.
We’ve raised multiple rounds of funding, completed early product validation, and shipped to users in 50+ countries, with backing and use cases involving top Olympic athletes. We’re in a fast iteration and growth phase, expanding into more sports and movement scenarios.
Our team is deeply technical, with early team members from MIT, Harvard, and leading companies.

What You’ll Work On

Build robust, interpretable signals and metrics from long-horizon sensor time-series data (e.g., baseline, variability, anomaly deviation, state changes)
Develop modeling approaches across users and scenarios (e.g., stratification, normalization, drift detection, anomaly detection) to improve consistency and reliability
Build personalization models that learn an individual’s normal pattern and detect early risk signals or performance changes
Design evaluation and regression pipelines to measure model/rule improvements and monitor performance over time
Collaborate with engineering to productionize algorithms: define interfaces/fields, support versioning, and enable fast iteration

What We’re Looking For

3–7 years of experience in applied data science / machine learning / algorithm development
Strong experience with at least one of the following: time-series data, sensor data, telemetry, or large-scale behavioral/log data
Solid Python + statistics / ML fundamentals, with the ability to handle real-world issues: noise, missingness, distribution shift, weak labels, etc.
Experience with (or strong interest in) heterogeneity and longitudinal / sequential modeling
Strong engineering mindset: writes maintainable code, values evaluation rigor, reproducibility, and iteration speed