Software Engineer

Preseen; United States; Remote; USD 110,000 - 180,000; Full time

Benefits offered by this job

Equity 0.1%–0.3%

Health benefits

Flexible time off

Job summary

Preseen is hiring remote engineers to help turn AI forecasting advances into a reliable product. We seek full‑stack generalists, with strengths in agents, data, backend infrastructure, or customer interfaces. Our stack includes Python, TypeScript, React, and Next.js.

You will work on measuring forecast quality, connecting agents to data sources, and building the product interfaces and back‑end systems that empower customers to ask useful questions and act on forecasts.

Qualifications

• Experience shipping web applications.
• Ability to reason about complex systems and uncertainty.

Responsibilities

• Measure and improve forecast quality through evaluations and experiments.
• Connect agents to data sources, decide what to store and refresh, and manage evidence in context.
• Build interfaces and backend systems that help customers query forecasts and act on them.

Skills

Web apps

Tools

Python

TypeScript

React

Next.js

Job description

Fully remote · Working hours that overlap with U.S. time zones

Every important decision contains a forecast.

Should a government impose sanctions? Should a company reroute its supply chain? Each decision depends on beliefs about what happens next. Preseen builds AI systems that help people make those decisions with better evidence and clearer estimates of uncertainty.

Our agents gather information, analyze it, and produce forecasts that update as the world changes. We measure their performance against what actually happens. Our systems have won multiple forecasting tournaments on Metaculus, and we’ve raised over $8M. Our small team brings together engineers, AI researchers, expert forecasters, and former senior government officials.

We’re hiring engineers to help turn advances in AI forecasting into a reliable, useful product. Our greatest need is for full-stack generalists, but we also welcome engineers with particular strengths in agents and data, backend infrastructure, or customer-facing interfaces'.

Your work could include:

• Measuring and improving forecast quality. Build evaluations and experiments to compare approaches, improve prompts and tools, and learn faster from results.
• Connecting agents to the world. Bring in new data sources and decide what to store, when to refresh it, and which evidence belongs in an agent’s limited context.
• Building the product. Develop interfaces and backend systems that help customers ask useful questions, understand forecasts, and act on them.

We’d like to hear from you if:

• You’ve built and shipped web applications. Our stack includes Python, TypeScript, React, and Next.js. Experience with some of these is helpful.
• You reason well about complex systems and uncertainty. You can investigate AI behavior that varies between runs and use experiments to distinguish real improvements from noise. You understand why forecast quality takes many outcomes to judge: events assigned a 60% probability should happen about 60% of the time.
• You take ownership and work well with others. You can turn an open-ended problem into useful software, explain your decisions, and seek help when you need it.
• You use AI development tools thoughtfully. You move faster with them while taking responsibility for the design and quality of what you ship.

We offer competitive pay, 0.1%–0.3% equity, health benefits, and flexible time off. You’ll have substantial influence over what we build and how we build it. We take the work seriously and value a life outside it.

The role is fully remote, with regular working-hour overlap with U.S. time zones. We may meet in person once or twice a year for a few days.