I
€55K–€70K/yrRemoteFull-timeMid levelData Science
Apply on Himalayas ↗
Listed on Himalayas. You apply through their page. We never charge applicants.
Pay
€55K–€70K/yr
Published in the job listing
Prepare for this job
Free on every jobLikely interview questions
At a glance
- Education
- Master's
Skills mentioned
dbtdatabricksaccountingpythonsqltableau
About the role
At In Tandem, we build technology that helps families manage everyday routines and navigate life’s biggest transitions. Through our four brands—OurFamilyWizard, Cozi, FamilyWall, and Custody Navigator—we help families stay organized, communicate well, and foster healthy childhoods.
We believe technology should strengthen relationships and make daily coordination less complicated. Everything we create is designed to lighten the mental load, reduce conflict, and support families through big and small moments.
You won't start from scratch. Our dbt project is live on Databricks with a star-schema finance mart in production and documented in Unity Catalog. What's missing is the layer on top, and that layer is yours. We also run an internal AI agent that answers business questions in Slack off the semantic layer, so every metric you define and test becomes a question Finance can ask directly rather than queue.
If you want your work to make a real difference in the daily lives of parents and kids, In Tandem is the place where your impact will truly matter.
As an Analytics Engineer (Finance & Modeling), you'll own the metrics, reporting and reconciliation our financial decisions rest on, turning subscription and transaction data across OurFamilyWizard, Cozi, FamilyWall and Custody Navigator into numbers the business can defend. As that foundation becomes automatic, the work shifts toward forecasting and modeling.
As an Analytics Engineer (Finance & Modeling), you'll own the metrics, reporting and reconciliation our financial decisions rest on, turning subscription and transaction data across OurFamilyWizard, Cozi, FamilyWall and Custody Navigator into numbers the business can defend. As that foundation becomes automatic, the work shifts toward forecasting and modeling.
You won't start from scratch. Our dbt project is live on Databricks with a star-schema finance mart in production and documented in Unity Catalog. What's missing is the layer on top, and that layer is yours. We also run an internal AI agent that answers business questions in Slack off the semantic layer, so every metric you define and test becomes a question Finance can ask directly rather than queue.
What you'll accomplish:
- Governed metrics. Add the definitions the finance mart doesn't have yet (ARR, renewal rate, churn, CAC), versioned in dbt and documented in Unity Catalog, with the tests that make each one defensible, and wired into the agent so Finance can ask them directly.
- Reporting Finance can run without you. Move the recurring finance pack into marimo, retire the Tableau views behind it, and turn a monthly hand-build into something that runs on a schedule.
- Numbers that reconcile. Reconcile transactions across the app stores, payment processors, subscription systems and the ledger, with automated checks that alert rather than surface at close. Route discrepancies upstream with the evidence attached and turn each one into a test. You build the checks and escalate. Accounting owns the treatment.
- Forecasting for Finance. Cohort subscriber and revenue projection, and aggregate renewal and retention forecasting, feeding the CFO's forecast and board reporting. LTV by acquisition source, feeding paid spend allocation with Growth. Pricing and promo impact, modeled before we ship rather than measured after. Individual churn propensity and uplift follow once there's a retention motion ready to act on the scores.
- Models ship as scored tables in dbt and Unity Catalog, with tests, lineage and a definition the agent can answer from. Not as notebooks.
Who you are:
- Statistically rigorous. You know why a cohort curve is the wrong shape, what a model is leaking, and when a result is noise.
- Fluent in finance, and unwilling to let a number you can't explain go out the door. When two systems disagree, you find out why and tell the people upstream.
- An engineer who owns the number end to end, from a CFO's question through the model, the test, the report and the conversation that follows. A spreadsheet nobody can reproduce reads to you as a liability.
- AI-proficient in practice, not in theory. Claude Code, coding agents and agent SDKs are part of our daily loop, and we expect you to already work this way. A lot of what used to be a week of modeling work is now an afternoon, and we're hiring for people who have internalized that.
- Willing to wear several hats. This is a small team supporting four brands. Some weeks you're deep in a retention model, some weeks you're chasing an app store discrepancy or unblocking someone else's dashboard. The scope above is the center of the role, not a fence around it.
- Motivated by work that matters. Families rely on these products during real moments in their lives.
What you bring:
- 3+ years in analytics engineering, data science, BI engineering or advanced analytics.
- Statistical modeling shipped into a business process: forecasting, survival or cohort analysis, or propensity.
- Measuring the lift from a price change or a growth experiment. You don't need to have designed the test, but you need to be able to read the result honestly and say what it's worth.
- Subscription metrics you've built and defended to a finance stakeholder: ARR/MRR, renewal, churn, LTV/CAC, or cohort revenue.
- Strong SQL and real dimensional modeling. You've designed marts, not just queried them.
- Python and Git as daily habits.
Nice to have, not a dealbreaker:
- dbt, Databricks and Unity Catalog, or a comparable lakehouse.
- Reconciling transactional data to a financial system of record.
- Code-first notebook reporting (marimo, Hex, Streamlit, Quarto).
Originally posted on Himalayas