Staff ML Engineer – Adaptive Learning

About Brilliant

Brilliant's mission is to create a world of better problem solvers. We make games for learning in math, science, computer science, and data analysis, for iOS, Android, and web. On Brilliant, you learn by doing – there are no videos, everything is interactive. We get you hands-on, figuring things out on your own. We help learners develop intuition through interaction, build understanding through experimentation, and have fun.

We serve hundreds of thousands of paid subscribers, and we’re hoping you might be the right person to contribute to accelerating our footprint to millions of customers (and changed lives). In addition to what’s below, you can see all open roles and learn more about our team culture on our careers page.

We have always prioritized building a healthy business as the backbone of achieving our mission. We are default alive (will be profitable before needing to raise), don't over-hire, are growing new customers at an exciting pace (high double-digits year-over-year). Our investors are top-tier + mission aligned, and we’ve kept our valuations tethered to reality – we aren’t playing “catch up” like many others.

In our day-to-day, we value adventure, excellence, generosity, and candor. We are optimists in the face of uncertainty, we take pride in our work, we go the extra mile for each other, and we tell it like it is (the good and the bad). We’re all here to do the best work of our lives together, and have a lot of fun along the way. 

We believe that real-time collaboration and human connection are necessary ingredients in building a high-velocity, creatively-oriented consumer product. We maintain core hours (10am - 3pm Pacific) where everyone is online, regardless of timezone. Over half of us are located near our hubs in SF and NYC, and folks outside of those cities travel to attend team offsites once-per-quarter.

The Role

We’re looking for an exceptional builder who can turn messy learner‑interaction data into delightfully personalized lesson sequences. You’ll collaborate closely with data engineers, product managers, and content creators to craft recommendation systems that help every learner find the right challenge at the right moment. You’ll be a dedicated ML engineer building on a mature analytics foundation, not starting from zero.

You will bring deep recommender‑system know‑how, a pragmatic shipping mindset, and curiosity about emerging AI techniques (including the intersection of recommender systems and LLMs) to shape the future of Brilliant’s learning experience. You will be joining a passionate and experienced team of entrepreneurial-minded people who are working to make a meaningful impact on the world.

Be prepared to discuss a recommender system you’ve owned end‑to‑end (with details you can legally share).

• Design, train, and deploy adaptive‑learning recommender models, targeting a measurable lift in lesson completion and learner retention.
• Own the full ML lifecycle—from Snowflake queries and feature tables, through model training, to a low‑latency inference service.
• Evaluate and introduce the right technique for the job—two‑tower retrieval today, graph‑based or knowledge‑graph models tomorrow—balancing impact, complexity, and scale.
• Launch and analyze A/B experiments, iterate quickly, and communicate results and next steps to cross‑functional partners.
• Document best practices, raise the engineering bar, and mentor teammates eager to grow their ML craft.

• Have 5+  years of hands‑on experience shipping recommender or ranking systems that moved key product metrics.
• Have deep expertise in Python plus strong experience with a modern deep‑learning framework (PyTorch or TensorFlow) and cloud‑native deployment.
• Have built at least two of the following in production: two‑tower retrieval, deep cross‑feature ranking models, graph or knowledge‑graph recommenders, sequential/Transformer recommenders, or contextual bandits.
• Routinely design and read A/B tests, translating statistical lift into plain‑language insights for product and content teams.
• Thrive in a builder’s environment: you bias toward action, sweat the details, and are happy to own a problem end‑to‑end.
• Stay curious about the frontier where classical RecSys meets generative AI, and are eager to explore new techniques when they promise real learner value.

Compensation and Benefits

We use a systematic compensation framework: salary scales are set each year for each job vertical, managers level folks on their team, and those levels are mapped directly to our compensation scales. A location-based adjustment is applied outside of SF and NYC (typically 5-10%) - feel free to ask us about your location!

Given the systematic approach, we always make First and Best offers - there is no negotiation (for new hires nor our existing teammates). This ensures people are paid based on their expected contribution, not their negotiation skills.

We offer top-notch health care plans, with 100% of the premiums covered for medical, dental, and vision for employees. About 1/3 of our team are parents, and we provide generous parental leave + up to $1900/mo in dependent healthcare coverage.

We offer flexible PTO, with a norm of taking off about 6 weeks per year (including federal holidays). We also provide home office equipment, a professional development stipend, and free food at our offices.

Our CCPA Privacy Notice can be found here.