> Markdown version of [/jobs/ext/4766404-gtm-data-scientist](https://www.wearedevelopers.com/jobs/ext/4766404-gtm-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # GTM Data Scientist - **Company:** Intercom - **Location:** San Francisco, CA, United States (Remote available) - **Experience:** Starter - **Salary:** $147,000 - $176,000 - **Contract:** Internship / Graduate position - **Skills:** Airflow, Business Logic, Code Review, Github, Statistical Hypothesis Testing, Machine Learning, Standard Sql, Salesforce.Com, Snowflake, Model Validation, Pandas, Scikit Learn, Performance Monitor, Software Version Control, Data Pipelines - **Published:** November 18, 2026 - **Apply:** https://job-boards.greenhouse.io/intercom/jobs/8374568 ## About the Role * 1-3 years in data science, analytics engineering, ML engineering or a related quantitative role (internships and graduate research count) * Some experience with a model or data pipeline that ran in production, or a strong wish to learn * Exposure to GTM data (Growth, Marketing, Sales) in a B2B SaaS setting is a plus Technical skills * Strong SQL and Python for data analysis and modeling (pandas, scikit-learn or similar) * Solid foundations in statistics: distributions, hypothesis testing, model evaluation * Comfort with version control (GitHub) and code review * Familiarity with dbt, Snowflake, Airflow or Salesforce is a plus ## Description Fin's GTM Data Science team builds the machine learning behind our Growth, Marketing and Sales teams. Our models decide which leads become MQLs, how they're routed, what they're worth, which play a rep runs next, how marketing spend is attributed, and which customers are ready to expand. We already run many production models, including lead scoring, lead value, next best play, multi-touch attribution, marketing mix modeling and customer expansion. Each one is owned end to end: deployment, monitoring, adoption and measured business lift. Keeping these models accurate, trusted and improving is now as important as building new ones. We're hiring a Data Scientist to own the day-to-day health of these production systems and to help turn prototypes into production models. This is not a reporting role. You'll ship changes that go live, and you'll see the effect on the pipeline. If you want to work on applied ML in production, with real stakeholders and revenue on the line, this role is for you. What will I be doing? Own production models day to day * Keep production GTM models healthy: inbound lead scoring and value, account and play recommendations, and marketing attribution and mix models * Maintain the business logic around them: MQL routing tiers and exclusions, play eligibility rules and competitor overrides, and lead value tier bands * Monitor performance, calibration, drift and data quality, and fix issues at their root cause: a broken enrichment feed, a tier mix that shifts overnight * Write documentation and runbooks: how each model works, how to retrain it, and what to do when it breaks, so no model depends on one person You keep the models the business relies on healthy and trusted. Be the first point of contact for GTM stakeholders * Answer questions from Growth, Marketing, Marketing Ops, SDR, Sales, CS and RevOps leaders, such as why a lead didn't MQL or why an account got a play * Turn ad-hoc requests into scoped, prioritized work and well-tested changes * Size a change's impact before it ships * Build self-serve dashboards and monitoring so the same question doesn't come back You make it easy for GTM teams to trust and act on our data products. Ship new models into production * Productionize prototypes built by senior data scientists into reliable, scheduled pipelines * Integrate model outputs into the systems that act on them: Salesforce, ad platform conversion values, experimentation metrics and outbound play lists * Set up the monitoring, alerting and tests each new model needs * Measure lift after launch and report what changed You'll learn how a model goes from notebook to production, and own that path. What impact might I have Within 6-12 months, you might: * Be the named owner of one or more production models, end to end * Cut the time from a stakeholder question to a shipped fix from days to hours * Catch data-quality breaks before they cost us MQLs or pipeline * Ship at least one prototype into production and measure its lift ## About Intercom - **Size:** 1,001–5,000 employees - **HQ:** San Francisco, United States of America [Company profile](https://www.wearedevelopers.com/companies/4167-intercom)