Note: The job is a remote job and is open to candidates in USA. Cozi is a part of In Tandem, a company focused on building technology that helps families manage everyday routines. As a Product Data Scientist, you will use behavior analysis and predictive modeling to enhance app experiences and collaborate with Product and Finance teams to drive business strategy.
Responsibilities
- Ship the models that power personalized in-app experiences — predicting what each member is likely to want next based on real-time behavior, family context, and engagement history
- Detect meaningful patterns in product usage data — communication, scheduling, and activity signals — and build the models that turn those patterns into in-app suggestions, alerts, and guidance for families
- Help replace fixed-placement UI logic with algorithmic decisioning that adapts every time someone opens the app
- Build the churn, retention, and activation models that help Product understand which behaviors predict long-term value and which interventions are worth shipping
- Develop predictive segmentation, propensity, and uplift models, refreshed continuously through automated pipelines so the rest of the business is always acting on current signal
- Partner with Finance on cohort-level subscriber and revenue forecasting, LTV by acquisition source, and sensitivity analysis on the assumptions that matter most
- Build pricing, promo, and refund-risk models that quantify the financial impact of monetization decisions before we ship them
- Train, deploy, and monitor models in Databricks and AWS, using Claude Code as your primary engineering interface
- When agentic delivery makes sense, wrap models in managed agents that surface predictions into the workflows stakeholders already use
- Move the team from 'what happened' to 'what would happen if' — uplift modeling, geo-tests, and A/B analysis grounded in proper causal framing
- A modeler who thinks in patterns, hypotheses, and tests, and who owns problems end to end — from question framing through SQL, feature pipelines, modeling, serving, and stakeholder communication. You ship models real decisions depend on, not just notebooks and decks
- Pragmatic about ML — you know when a model is worth building and when a heuristic does the job, and you translate a CFO's question and a PM's question with the same fluency
- AI-first in how you build. Modern AI tooling — Claude Code, agent SDKs, coding agents — is part of your craft, and you operate at higher leverage because of it
- Motivated by work that matters. Families rely on these products during real moments in their lives
Skills
- 5+ years shipping predictive models in production, ideally across personalization, behavioral prediction, lifecycle ML, or subscription forecasting
- Strong Python and SQL
- Hands-on experience with Databricks and AWS
- Demonstrated success with subscription or consumer-app metrics — LTV, churn, retention, activation, forecasting
- Solid grasp of experimentation and causal methods (A/B, DiD, uplift, geo-tests, or similar)
- Production rigor — feature pipelines, model monitoring, retraining cadence
Benefits
- Medical: In Tandem pays 100% of the premium for employees AND 99% for all additional family members
- 401k: Up to a 4% match with immediate vesting
- Paid leave for all new parents
- Learning & Development stipend for employees
- Paid Time Off: 11 Holidays + Winter Break (3 Days) + Volunteer Time Off (1 Day) + Floating Holiday (1 Day)
- Personal Time Off: 15 days for 0-1 years of employment, 20 days 1-3 years of employment
- Supportive and flexible working environment – work from anywhere!
Company Overview