Software Engineer II, Agriculture

Onshore•New York, NY

11d•Onsite

About The Position

Onshore is seeking a Software Engineer II for their Agriculture team. This role involves building software that helps farms claim R&D tax credits by analyzing their research activities. The team focuses on creating data pipelines to process farm records, a qualification engine to determine eligible research expenses, and reports that withstand IRS review. Responsibilities include developing data ingestion pipelines, building connectors to farm systems, extending the qualification engine, encoding tax methodology into automated checks, and contributing to new incentive types like livestock and dairy deductions. The role requires a strong focus on correctness and clear technical communication.

Requirements

2+ years designing, shipping, and operating production software, from requirements through architecture, delivery, and post-launch iteration.
Strong proficiency in a typed language and in SQL. We work in TypeScript; deep experience elsewhere with real interest in switching counts.
You've worked with messy data at scale: OCR output, inconsistent spreadsheets, ten formats for the same document.
You care about correctness beyond what most product engineering demands. Our output is reviewed by tax professionals and filed with the IRS.
Clear, precise communication on technical topics, in writing and in person.
Able to work in person at our New York City office.

Nice To Haves

Experience building document extraction or structured-output LLM pipelines in production.
Experience with third-party integrations: OAuth flows, webhook reliability, rate limits, and the general art of depending on someone else's API.
Background in agtech, or you grew up around farming and know what a co-op statement looks like.
Experience with scientific, trial, or geospatial data.

Responsibilities

Build ingestion pipelines that turn messy real-world records into structured, computable data: scanned invoices, co-op statements, tax schedules, agronomy reports
Build connectors to the systems farms already use, so document chasing stops being the bottleneck of every study
Extend the qualification engine: per-variety determinations with rationale that feeds both the computation and the report narrative
Encode expert tax methodology into automated report checks, so reviewer knowledge compounds instead of living in one person's head
Help build the methodology for new incentive types: livestock, dairy, and soil fertility deductions
Write the evals for what you ship; a wrong number here is a wrong number on a federal filing