Senior/Lead Software Engineer, Applied Experiments
Confirmed on the employer's own hiring board on Oct 20, 2026. First seen by Alion on Oct 16, 2026. Open Government Products scores B on the Alion truth index.
Who We Are
Open Government Products (OGP) operates like a startup within the government. We are a fast-paced, dynamic team of engineers, designers, and product managers dedicated to public good. We autonomously identify opportunities to solve public problems using tech, rapidly prototype, and launch solutions, from citizen apps to automating public agency operations. Embracing an open-sourced and flat hierarchy, we cut through bureaucracy to focus on real-world, user-centric innovations that drive change.
As part of the Applied Experiments Division (AED) within OGP, we take on the most ambiguous open problems and run experiments to help inform better decisions in government. Some past examples like ERP X, Armoury, Maps, Pair and ParkingSG are just a few examples of how we've made a difference in the lives of citizens. You can learn more about our products on our website.
About the role
We're looking for a Senior/Lead Software Engineer who can take an unclear problem and turn it into something worth testing. You'll work closely in a small team of 1-2 other engineers and/or product managers and designers, and take ownership of experiments from the first user conversation through to launch and learning.
You'll have room to influence what we build, and how we build it. That comes with responsibility: finding the next useful step, making trade-offs, following through, and changing course when the evidence calls for it. These experiments eventually become policy or products that influence people’s lives, or learnings that will inform better policymaking or product building down the road.
What you'll do
Turn vague, ambiguous problems into actionable experiments. Talk to users, identify the riskiest assumption or problems, scope the smallest build that tests it, and deliver in weeks. Know when more engineering helps, when it doesn’t, and define what you need to learn from it.
Build and ship across the stack. Take an experiment through implementation, testing and deployment. Keep scope focused, make sensible technical choices, and apply the care needed for real users to depend on it.
Own the learning. Combine user feedback with product data to assess what happened. Recommend whether to improve, expand, change direction or stop, and explain your reasoning. Work with PMs, designers and policy officers to make sure the results reach the right people to inform decision making.
Follow through on what’s been built. Keep live experiments safe and reliable, then help hand them over, merge them or wind them down responsibly once they’ve answered the question we set out to test.
What we’re looking for
Strong, hands-on engineering. You build end to end, quickly and simply, at a standard that fits the stage; Even prototypes gets clear data boundaries and basic security.
Open-ended ownership and initiative. You take responsibility for a problem or domain, not just a task. When the way forward isn’t clear, you’re able to ask good questions, work with the team to shape an approach, and follow through without waiting for top-down directives.
Problem-first, not solution-first. You’re invested in solving the problem, not building a solution, and you’re comfortable changing course when the evidence says so.
Comfort with ambiguity, adapting and learning. You’re comfortable starting without all the answers, testing your assumptions and changing direction based on what you learn. Ruling out an approach is useful progress when it helps the team decide what to do next.
Collaboration across disciplines. You work alongside PMs, designers and policy officers to shape problems and test ideas. You explain technical possibilities and trade-offs clearly, and draw on others’ expertise to make better decisions together.
Experience using AI effectively. You’ve used AI, whether through agents or by embedding it in what you build, to help you design, build, test or operate software. You can explain where it fell short and how you checked its output, and you stay accountable for the quality of what ships.
A genuine drive to do public good. You care about outcomes for stakeholders and citizens, you’re accountable for the resources invested in experiments, and you’re motivated by overall public good impact.
What matters less than you'd expect
Years of experience. We care about what you’ve owned, the decisions you made and what you learned. Time served isn’t an accurate proxy for any of that.
Related degree or formal credentials. CS degrees, ML backgrounds and policy qualifications are all welcome, none required.
Specific language, framework or stack. Fundamentals and the ability to learn fast matter more.
Specialist AI infrastructure experience. Demonstrated, practical use of AI to build and operate software matters more than depth in agent or evaluation tooling.
Government or policy experience. We’ll teach you how decisions move through government, and how to navigate the bureaucracy, but we can’t teach you to take initiative.
Having worked on long-lived products. Experience shipping and maintaining production software helps, but a polished product isn’t the goal here, a well-answered question is.