As a Forward-Deployed Data Scientist, you’ll partner with our CEO, customers and engineering team to understand customer datasets, frame critical finance workflows and automate them with our AI agents. You’ll transform messy ERP data into reliable data models, work with our AI agents to generate insights and narratives, and help customers adopt agent-powered finance processes.

You’ll work directly with the CEO, CTO, Product, and Design in a highly collaborative, customer-facing role.

Key Responsibilities

Review customer financial datasets, reconcile gaps, and ensure 100% coverage across our data models.
Ship dbt models, seeds for COA mapping/overrides, tests & docs; design for backfills, SCDs, idempotency, and data quality SLAs.
Build clear analyses on datasets to surface unusual patterns, priorities and actions.
Iterate on prompts and workflows so AI agents perform increasingly complex finance tasks (e.g., MD&A drafts grounded in GL detail).
Run customer discovery: map current processes, identify automation opportunities, and translate them into requirements.
Partner with engineering to train/evaluate agents; define success metrics and benchmarks.
Work with engineering team to train AI agents on various use cases.
Monitor agent-generated data signals to detect anomalies and key drivers.
Curate and present findings to finance leaders; coach customer teams on best practices.
Provide product feedback and help shape features that deliver measurable time savings and accuracy.

Qualifications

A great candidate will possess the following attributes:

3-6 years of experience
SQL expert and experience with dbt (models/seeds/tests)
Firm understanding of modern data warehouses (e.g., BigQuery)
Accounting/Corporate finance fluency incl. multi-entity consolidation, GL/account mapping.
Firm understanding of Excel/Sheets modeling (driver-based forecasts, scenario analysis).
Practical financial data experience plus familiarity with CRM/HRIS/commerce/payments data.
Customer-facing communication: discovery, documentation, and storytelling.
Comfort in fast, ambiguous 0→1 environments; bias to ship.

Nice-to-haves