Minerva Forward Deployed Engineer at Minerva building custom data ingestion, pipelines, and models for complex customer accounts. Collaborate with CTO and data science team.
Responsibilities
Our solutions engineers get standard accounts live on our product. Some accounts need more than the product does today. A pro sports team with ten years of ticketing data in three systems. A brand that needs a model we haven't built yet. First-party data that needs real work before anyone can trust it.
Those accounts get you. You're a data scientist who ships production pipelines, or a data engineer who builds models, embedded with our most complex customers. You build the custom onboarding, pipelines and models that make Minerva work in their world, then bring what you learned home so the next customer gets it out of the box.
You work closely with our CTO and data science team, alongside each account's AE, forward deployed marketer and solutions engineer.
Get the data ready. Build custom ingestion and pipelines for accounts whose data doesn't fit standard onboarding, and make the call on whether the data is right before anyone builds on it.
Build custom models. Propensity, lookalike and scoring models trained on a customer's first-party data and our consumer graph, when the standard models aren't enough.
Run scoring in production for your accounts, and own its accuracy.
Build the net-new. When an expansion needs something we don't have, scope it with the forward deployed marketer and build it.
Make the one-offs standard. Turn custom pipelines and models into patterns our data platform team can productize.
Be the technical voice in the room when a customer's data team wants to know how it actually works.
4+ years as a data scientist or data engineer shipping to production in a data-heavy setting: models that ran on real data, pipelines other people depended on.
Strong Python and SQL, experience with dbt and an orchestrator like Dagster or Airflow, and a cloud warehouse (Snowflake, BigQuery or Databricks).
You've built predictive models such as propensity, LTV, lookalike or lead scoring, and you know how to tell a customer when their data won't support one.
Qualification
Consumer dataExperiment designForward deployedSportsStandard onboardingA research roleWHAT SUCCESS LOOKS LIKEFirst 90 days
Preferred
Consumer data, identity graphs, adtech or martech.
Experiment design: holdouts and lift measurement.
Forward deployed, solutions or consulting experience at a data company.
Sports, financial services or travel, where our enterprise motion is strongest.
Standard onboarding. Our solutions engineers get standard accounts live on our product. You come in when an account needs something built, before or after the deal closes.
A research role. Everything you build runs for a paying customer.
WHAT SUCCESS LOOKS LIKE
First 90 days: you've taken one complex account from raw data to production scoring. Your data-readiness call is written down, and the customer's data team is on your side.
