Data Scientist

Applied intelligence · Remote · Full-time

The capabilities Frontier delivers rest on models — forecasting, driver analysis, scenario comparison, anomaly detection — built on live operational data out of a company’s ERP. This role builds them, and then takes them the whole way: from the notebook that proves the idea to a governed capability running in the product, with its assumptions visible to the finance team relying on it.

What you’ll do

Build the models behind the capabilities: demand and financial forecasting, driver analysis, scenario comparison, anomaly detection.
Work on live operational data — ledger, transactions, inventory positions, pipeline — with everything that implies about quality, history and gaps.
Take a model from notebook to a governed capability in the product: evaluated, monitored, and honest about what it assumes.
Make the reasoning inspectable, so a finance team can examine an assumption instead of trusting an output.
Work with the forward deployed engineer on what the customer’s problem actually requires, before choosing the method.

What you bring

Statistics and machine learning foundations you can defend — you know why a method is the right one for this problem.
Python, and the data stack around it, at production quality.
Time-series and forecasting experience on data with real seasonality, promotions and holes in it.
Driver and causal analysis: separating what moved the number from what merely moved alongside it.
You can explain a model’s assumptions to a CFO, in their language, without giving up what is true about it.
You care that the model runs in production. A result that never leaves the notebook is not a result.

Working context

Remote, full-time. Anywhere in the world, US preferred. Salary plus equity in the company.
Reports to the founder, and works with engineering to put every model into the product.
Early-stage — you define what applied intelligence means at Frontier.

Apply for this role →