Evaluation Engineer at Elicit building auto-evaluation systems for an AI research platform. Focus on infrastructure speed, interfaces for ML engineers and product managers, and statistical rigor for pharma decision-making assessments.
Elicit is an AI research platform that uses language models to help researchers figure out what's true and make better decisions, starting with common research tasks like literature review.
What we're aiming for:
Elicit radically increases the amount of good reasoning in the world.
For experts, Elicit pushes the frontier forward.
For non-experts, Elicit makes good reasoning more affordable. People who don't have the tools, expertise, time, or mental energy to make well-reasoned decisions on their own can do so with Elicit.
Elicit is a scalable ML system based on human-understandable task decompositions, with supervision of process, not outcomes. This expands our collective understanding of safe AGI architectures.
Visit our Twitter to learn more about how Elicit is helping researchers and making progress on our mission.
Some orgs build evals to warn us about dangerous capabilities. Others build evals to understand trends and predict future developments. Yet others build evals to hill-climb towards models that users will like more.
At Elicit, we're focused on something different—we want to understand, and hill-climb towards, models that help us make better decisions.
This is tougher than "what will users like better"—it's hard to evaluate decision support, and users' knee-jerk reactions may not align with what actually helps for decision-making. Because it's hard, and because the sales pitch is more complicated, there aren't many doing this well. If we nail this, we have a unique opportunity to push AI toward helping us make better decisions, both within Elicit and beyond.
We need someone to own the technical foundation of our auto-evaluation systems. Our evals are currently much slower than they need to be, and our interfaces aren't optimized for the diverse set of people who need to use them—ML engineers iterating on models, product managers monitoring quality, and customers assessing trust in results.
The right person for this role won't just build infrastructure. You'll think deeply about what it actually means for Elicit to help with decision-making in pharma and encode that understanding into our evaluation systems.
The core auto-eval platform
You'll build a comprehensive system that runs fast, is easy to use, and supports quickly building new evals:
Speed: You’ll build a lightning-fast basic evals infrastructure that schedules tasks to introduce practically no latency; and then you’ll figure out clever ways to solve the fundamental sources of latency (building a version of Elicit, running it on a query, and evaluating it using LMs)
Interfaces: ML engineers need evals to kick off automatically on relevant commits, with results they can see at a glance and drill into. Product managers need dashboards showing performance over time and what's going wrong in production.
Architecture: Your code must be well-architected so other team members and ML engineers can understand and build on it. An engineer starting on a new feature should be able to quickly add examples and run an eval.
Ensuring evaluations are accurate and reliable
We need to evaluate how well Elicit actually helps with decision-making in pharma, not just measure what's easy to measure. This requires encoding real knowledge about how pharma customers make decisions (for example, choosing appropriate gold standards).
You'll provide appropriate statistical tests and confidence intervals so we can trust our results.
In a typical month, expect to spend:
60% working on the core eval platform
15% working closely with the evals team to build and improve specific evals (e.g., an eval of our paper search within our systematic review flow)
10% mentoring our evals engineering intern
The rest on learning how people interact with the eval system so you can make it work better for them, and understanding what our users want from Elicit so evals measure what matters
Requirements
At least 3 years of experience as a professional software engineer, with demonstrated experience building complex backend systems (e.g., backend for a complex website, data pipelines, etc.)
Aptitude and interest in evaluating how Elicit helps with pharma decision-making. There's no particular experience you must have, but we'll evaluate your aptitude.
Will make you more competitive for the role
Knowledge of statistics (for e.g. calculating power and credence intervals for evals)
Experience with advanced Python (asyncio/trio and parallel processing strategies)
Front-end experience and strong UX sensibility (you'll be building dashboards). TypeScript experience is a plus.
Experience building developer tools (ML engineers are one of your most important clients)
Previous experience as a data engineer or working on AI infrastructure
Knowledge of pharma/biomed
Experience evaluating ML systems
Experience building language-model-based systems (helps with understanding Elicit and how to evaluate it)
This is a diverse list of nice-to-haves. We expect the candidate we select to have some, but not all, of these. Other team members can fill in for skills you lack.
We have a lovely office in Oakland, CA, but we’re flexible about where you work. You’re welcome to work remotely, from our Oakland headquarters, or in a hybrid setup. The only in-person requirement is attending our quarterly team retreats, typically held on the west coast.
In addition to working on important problems as part of a productive and positive team, we also offer great benefits (with some variation based on location):
Flexible work environment: work from our office in Oakland or remotely with time zone overlap (between GMT and GMT-8), as long as you can travel for in-person retreats and coworking events
Fully covered health, dental, vision, and life insurance for you, generous coverage for the rest of your family
Flexible vacation policy, with a minimum recommendation of 20 days/year + company holidays
401K with a 6% employer match
A new Mac + $1,000 budget to set up your workstation or home office in your first year, then $500 every year thereafter
$1,000 quarterly AI Experimentation & Learning budget, so you can freely experiment with new AI tools, take courses, purchase educational resources, or attend AI-focused conferences and events
A team administrative assistant who can help you with personal and work tasks
For all roles at Elicit, we use a data-backed compensation framework to keep salaries market-competitive, equitable, and simple to understand. For this role, we target starting ranges of:
Career (L3): $140-170k + equity
Senior (L4): $165-200k + equity