Model Behavior Engineer, Reasoning and Decision Making
Elicit · Oakland, CA (or remote within US timezones)
About this role
## About Elicit Elicit helps research and business teams investigate hard questions and make high-stakes decisions grounded in evidence. Our ambition is to make high-quality reasoning under uncertainty available at a much greater scale in contexts where stakes are high and feedback is sparse. We’re building AI systems whose work people can inspect, question, and improve—so increasingly capable AI also makes us more wise. ## Why we’re hiring Some of the most valuable questions people face are also the hardest to answer well—e.g., which research program to advance, what evidence would change priorities, and which uncertainty is worth resolving before committing to a multi-year, $100M trial. Today’s models can contribute, but they often: - miss crucial considerations, - make assumptions they should have checked, or - reverse recommendations too readily when new information arrives. Because models can work more autonomously, their process and reasoning can be poorly legible—so decision makers are wary to trust outputs. We’re looking for someone to own this work: understand where our systems fall short, develop ways to evaluate them, and build improvements that help users make better decisions. ## About the role You’ll join our engineering team and report to an Engineering Manager. You’ll work on a cross-functional engineering team collaborating with engineers, product, design, and domain experts—from investigation and experimental design through implementation and use in the product. You’ll design agent environments and how agents represent problems, gather evidence, use tools, retain context, revise conclusions, and coordinate with each other. You’ll also make engineering decisions that let us inspect processes, test changes, and build on what works. Most initial work will be built around existing frontier models. Training or finetuning may be useful for particular problems, but the direction will follow what helps users reason better. ## What you’ll own - **Understand the decisions we’re helping people make** - Work through real cases with users and domain experts - Identify objectives, constraints, available evidence, and key unknowns - Learn enough about unfamiliar problems to recognize when an answer misses the point - **Define and measure good reasoning** - Build evaluations where success is initially underspecified - Combine close inspection of examples, expert judgment, and quantitative analysis - Determine whether better scores reflect meaningful improvements that generalize - **Improve the system’s behavior** - Turn observed failures into testable hypotheses - Experiment with decomposition, context, tools, and ways for agents to seek information or check their work - Follow promising results through to user-facing changes - **Design reliable software** - Choose useful abstractions, data structures, and component boundaries for reasoning workflows and evaluations - Make tradeoffs among simplicity, flexibility, reliability, latency, and cost - Build systems whose failures you and your teammates can diagnose - **Own the quality of what you ship** - Write maintainable code, participate in design/code reviews, and test/monitor changes - Integrate systems into the product and improve based on real usage ## Example projects - Build and evaluate a workflow to help research teams compare protein targets for drug development—aligned to goals and decision frameworks, legible against stress-tested theo
Listing freshness
CronJobs last confirmed this listing 58m ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.