Software Engineer, Data Platform

About the role

WHY NUMERIC

Every business runs on accounting, but the systems underneath it were built for a different era. Today’s ERPs don’t store a company’s financial data. They store a diluted copy of it. Every contract, lease, and invoice gets flattened into a journal entry, and the context that made it meaningful is thrown away. That’s why accounting teams spend every month re-checking numbers, stitching together spreadsheets, and waiting on engineering, and why AI struggles to do real accounting work. Accounting doesn’t have a reporting problem. It has a data problem.

Enter Numeric: the data platform finance teams run on. We started with the month-end close, where the data problem hurts most, and hundreds of finance teams now run their close on Numeric. That work led us to rebuild the system of record itself. In September 2026 we launched the Financial Data Platform (FDP), a platform built for the agent era. Instead of a journal entry, the real business object enters the system and the accounting is derived from it. Every number is traceable to its source, reporting works on any dimension, and accountants and AI agents finally have enough context to work together. While others bolted AI features onto the old data model, we built a new one.

We have a growing base of customers who love our product, including Anthropic, Plaid, Brex, Betterment, Wealthfront, and Asana. We’ve raised $89M from top-tier investors including IVP, Menlo Ventures, Founders Fund, and 8VC, along with founders and executives from Ramp, Segment, and other category-defining companies. We’re in-person, with teams in New York, San Francisco, and London, and we just shipped the product that will define our next decade. There’s never been a better time to join.

About the role

Numeric’s product is, at bottom, data: your accounting history, your ability to pass an audit, your capacity to catch mistakes before your auditors do. The Data Infrastructure team builds the financial data plane that makes that possible — the lakehouse that ingests raw data from ERPs, banks, payment processors, and customer warehouses, transforms it into typed canonical tables, and serves it to our accounting modules, reporting engine, and AI agents. We’re mid-way through a deliberate architectural bet (ClickHouse-backed lakehouse, SQLMesh, per-tenant isolation) and it is the layer we expect to carry the most new investment this year. Every engineer on the team owns a slice of the architecture end to end.

What you might work on

- Ingestion at scale: incremental, watermark-based extraction from API sources (NetSuite, Stripe, Brex, Ramp, HubSpot…), file-based S3 inboxes, and Postgres → ClickHouse replication — and making each new source cheap to add.

- Data contracts & typed source data: stop downstream consumers from touching untyped raw JSON; enforce schemas at the boundary and make datasets discoverable.

- Data Lakehouse schema management: a single declare → codegen → apply pipeline so schema intent is codified, reviewable, and applied consistently across tenants.