Worked on applied AI or ML problems where you had to structure your thinking, test hypotheses, and draw conclusions from messy or incomplete data - doesn't need to be formal research, but you should be able to show how you approached ambiguity methodically,
Had some exposure to evaluating AI or ML systems - whether that's writing test suites, tracking model performance metrics, or thinking critically about why a model fails in certain scenarios. We're not looking for a fully-formed eval framework, but you should understand why this matters,
Worked in or alongside production AI systems - enough to understand the gap between a prototype that works in a notebook and something customers actually depend on. Direct production experience is a plus, but strong intuition from adjacent work counts,
Communicated technical findings clearly to non-technical audiences - whether that's a product manager, a stakeholder, or a customer. You don't need a track record of boardroom presentations, but you should be comfortable translating what you found into what it means,
Written Python comfortably across the AI/ML stack - data processing, model evaluation, experimentation. You don't need to be a research ML expert, but you should feel at home in the tools,
Been part of cross-functional teams where research or analysis fed into real product or engineering decisions - comfortable sitting at the boundary between exploration and execution, even if you're still developing confidence there,
If this opportunity resonates with you, we encourage you to apply
What the job involves
The Research Engine sits at the intersection of exploration and evidence-based decision-making. We ensure that major technical investments are de-risked through rigorous research before Engineering scales them into production. Our work shapes what gets built, how it's built, and whether it's worth building at all,
As a Research Engineer, you'll bridge Product Management and Engineering by investigating emerging AI capabilities, validating technical approaches, and producing evidence that guides strategic product decisions,
You'll design evaluation frameworks that measure agent performance consistently, run structured experiments to test competing hypotheses, and translate complex findings into clear recommendations that influence roadmap priorities and implementation choices,
This isn't abstract research; it's applied investigation with direct commercial impact. Your work accelerates Engineering delivery by reducing uncertainty upfront, strengthens product quality by identifying failure modes early, and builds institutional knowledge about what works in one of AI's most complex application domains: specialty insurance pricing and underwriting,
Design and maintain evaluation frameworks that enable consistent, automated measurement of AI agent performance across hundreds of insurance scenarios, creating the testing infrastructure that catches regressions before customers do,
Conduct structured technology assessments that evaluate emerging AI capabilities against specific product needs, producing technical briefs and competitive analyses that inform roadmap decisions worth millions in engineering investment,
Run disciplined experiments that decompose ambiguous research questions into testable hypotheses, using statistical rigour and practical engineering insight to generate reliable findings under tight delivery timelines,
Build comprehensive test suites covering actuarial edge cases, underwriting workflows, and pricing logic that stress-test AI systems in ways that mirror real customer usage, improving agent reliability by 40%+ through targeted improvements,
Translate research findings into engineering action by identifying performance gaps, recommending technical approaches, and working alongside Product and Engineering to turn insights into shipped capabilities customers trust,
Maintain domain fluency in insurance by staying current with actuarial concepts, underwriting practices, and pricing methodologies, ensuring research outputs respect the complexity of the industry we're transforming