Data Science and Business Insights Lead

Job Title: Data Science and Business Insights Lead Department: Product Reports to: COO About Luna Joy Luna Joy is a leading provider of mental health services, specializing in supporting women through their mental wellness journey. We offer holistic therapy and medication management to help women overcome challenges like postpartum mental health issues. Our dedicated team of professionals works in a collaborative, autonomous, and remote environment, ensuring comprehensive care for our clients. We are rapidly expanding and committed to improving access to quality mental health services across the U.S. Job Overview LunaJoy is a virtual behavioral health practice. Nearly every operating number the company runs on- provider utilization, payer mix, lead and referral conversion, clinician capacity by state, and weekly leadership scorecards- is produced by one dbt project on BigQuery. This role owns that project. This is a sole-owner seat. You will be the only person whose full-time job is the data stack: you set the modeling standards, you triage the broken sync at 8 am, you decide what a metric means, and you are the person Finance, Clinical Ops, Growth, and the exec team come to when they need a number they can defend. There is no data team above you to escalate architecture decisions to, and no analyst below you to hand ad-hoc queries off to. That autonomy is the appeal of the job and also its main demand; we are hiring for judgment as much as for SQL.

The role

is open because the current owner is departing. Expect a documented handover, not a cold start: the project carries a maintained CLAUDE.md , a DATA_ MODELS.md data-model reference with a grain and terminology glossary, an Electronic Medical Record source-system data dictionary, declared exposures, and dbt tests. As part of the Luna Joy team, you will play an essential role in directly contributing to supporting the mental health journey of one million women with purpose-driven innovation, unwavering compassion, and transformative care. You will need the speed to adapt quickly and efficiently, as well as the drive to be 1% better every day, constantly striving for improvement. Additionally, you will demonstrate compassion and contribute to a positive work environment by giving and getting joy in your daily interactions. These values will enable you to make effective contributions to key areas of focus and responsibility. Key Responsibilities: Data Modeling (dbt + BigQuery) Maintain and extend a production dbt project of roughly 220 models spanning appointments, patients, charges and billing, providers, referrals, lead and outreach conversion, capacity, and clinical measures Continue the project's layering convention (staging → int → mart) for new work while keeping a large set of established flat dashboards/models stable and correct Own the semantics of the hard domain logic, multi-payer insurance billing sequences, fee-schedule matching with its fallback hierarchy, patient active/churn state, care-model classification, provider utilization and capacity Keep tests, documentation, and declared exposures current so that a change to an upstream model reveals what it will break before it breaks it Ingestion and Pipelines Own the health of every source feeding the warehouse: the AMD practice-management system (ODBC), the internal warehouse, GoHighLevel CRM, Salesforce/Verifiable credentialing data, Airbyte-loaded external files, payroll, and GA4 Diagnose and resolve sync failures, schema drift, source-freshness gaps, and the quirks that come with incrementally-synced upstream systems Maintain the Python build scheduler and the scripts that sit outside dbt Reporting and BI Delivery Produce the recurring reporting layer, roughly 30 scheduled reports, including weekly analytics and referral reporting, monthly performance and forecasting, capacity and availability reports, and external partner-facing extracts Maintain the dashboards and Google Sheets / Connected Sheets that leadership and operators actually open, and keep the exposure layer honest about what depends on what Improve how numbers are presented, not just computed; a report nobody trusts or can read is not delivered Analysis and Stakeholder Partnership Serve as the analytics partner to four groups: Finance / RCM : charges, payments, adjustments, fee schedules, provider payroll, monthly performance, profitability Clinical Ops/provider network : utilization and fill rates, capacity by state, hiring and enrollment pipeline, credentialing coverage, documentation and charge timeliness, quality metrics Growth/partnerships : lead and referral conversion, source mix, time-to-conversion, care-partner and B2B pipeline, outreach targeting Exec leadership : scorecards, weekly and monthly leadership reporting, and the ad-hoc question that arrives an hour before the board call Turn vague business questions into defensible answers, and push back when the question is the wrong one Data Governance and Trust Handle protected health information responsibly and keep PHI out of places it shouldn't be Maintain test-data exclusion rules, business-rule definitions, and metric definitions that keep numbers consistent across reports Serve as the tie-breaker when two dashboards disagree.

Qualifications

Expert SQL: not "comfortable with joins," but window functions, complex CTE pipelines, incremental logic, and the ability to reason about grain and fan-out in a 15-CTE model you did not write Production dbt experience: you have owned or substantially built a dbt project- models, tests, sources, exposures, macros, materialization strategy, and the discipline of refactoring without breaking downstream consumers Fluent in Google BigQuery, including its performance and cost characteristics Demonstrated end-to-end ownership of an analytics stack, including the parts that aren't modeling: broken pipelines, undocumented source systems, and stakeholders who need an answer today Strong written communication: in a remo