Founding Engineer, Understanding Pipeline

Plotwell Inc.; San Francisco, Northern (CA, KY); Hybrid; USD 140,000 - 210,000; Full time

Job summary

Plotwell Inc. in San Francisco is seeking an early engineer to help turn data into trustworthy dashboards. You will work on backend and data systems that others rely on, shaping how we parse inputs and generate figures from real rows rather than estimates.

The role emphasizes building robust type inference, grounding checks, and efficient routing to reduce cost without sacrificing quality. You will collaborate with a founder, undergo a paid work session, and progress through a quick, transparent

Qualifications

• 5+ years building backend or data systems that other people depend on.
• You have worked with language models in production and have opinions about evaluation that come from experience.
• You write TypeScript comfortably, or will within a fortnight.
• You care about the difference between a demo and a thing an operator trusts on a Monday.

Responsibilities

• Make parsing survive real inputs: mixed date formats, currency symbols, merged headers.
• Own column type inference and the statistics behind every chart and outlier.
• Keep the grounding check honest: any takeaway that cites a number we cannot trace to the data gets dropped, and there is a test for it.
• Cut the cost per dashboard without cutting quality, which mostly means routing better rather than prompting cleverer.

Job description

Plotwell turns a spreadsheet into a dashboard and plain takeaways. The understanding pipeline is the product: reading a file a person made by hand, inferring what each column means, deciding which charts fit, and computing every figure from the real rows so the model never types a number onto a chart.

You would be an early engineer. The work is not glamorous and it decides whether the product is trustworthy. A dashboard that quietly invents a figure is worse than one that says less.

What you would own

• Make parsing survive real inputs: mixed date formats, currency symbols, merged headers, a stray total row at the bottom.
• Own column type inference and the statistics behind every chart and outlier.
• Keep the grounding check honest: any takeaway that cites a number we cannot trace to the data gets dropped, and there is a test for it.
• Cut the cost per dashboard without cutting quality, which mostly means routing better rather than prompting cleverer.

What we are looking for

• Five or more years building backend or data systems that other people depend on.
• You have worked with language models in production and have opinions about evaluation that came from being burned.
• You write TypeScript comfortably, or you will within a fortnight.
• You care about the difference between a demo and a thing an operator trusts on a Monday.

How hiring runs

A thirty minute call with a founder, then a paid work session on a real problem from our backlog, three to four hours, then conversations with two people you would work alongside. Four steps, usually inside two weeks, and we tell you where you stand at every one.