Most AI is frozen in place - it doesn't adapt to the world. We think that's backwards. Our mandate is to build efficient intelligence that evolves in real-time. Our vision is AI systems that are flexible, personalized, and accessible to everyone. We believe efficiency is what makes this possible - it's how we expand access and ensure innovation benefits the many, not the few. We believe in talent density: bringing together the best and most driven individuals to push the boundaries of continual adaptation. We're looking for builders and creative thinkers ready to shape the next era of intelligence.
Sweat the details. Technical excellence requires obsessing over every detail. We co-design serving, algorithms, and interface as one system to maximize efficiency and enable real-time adaptation.
Move with conviction. Extraordinary results require extraordinary effort. We operate with urgency as a unified team, concentrating resources on a few high-conviction bets where research and impact intersect.
Metrics that matter. The most intelligent system will increasingly be defined by building an algorithm that can interact with the world. Research ideas are tested through working products. If it doesn't improve what users can do, we question whether it matters. We share what moves the field; we don't optimize for paper count.
This is a research role which is focused on real world impact. We care about building our frontier innovation around efficiency, gradient free exploration, real time learning and interface design. A revolution is underway where the cost of generating synthetic data is now low enough that we can treat the data space as malleable and something which can be optimized. We can steer synthetic data towards desirable properties and make previously invisible worlds with limited data coverage more visible. The most intelligent system will increasingly be defined by building an algorithm that can interact with the world. This means for the first time researchers who care about intelligence need also be obsessed with how a model interacts. If any of these statements resonate with you, we would like to hear from you.
Innovation: innovate on our product to co-design algorithms that react real-time to product signal and feedback. Design new ways of giving feedback that drive better algorithms.
Cross-Stack Optimization: Collaborate across software, hardware, and algorithmic domains to achieve system-wide efficiency gains
Measure what matters: We believe strongly that the ultimate signal of value is whether we have real world impact. Our current algorithms are capable of interacting with the world which is why product matters so much.
A PhD or equivalent research experience in a computer science field
4-5+ years of industry experience
Deep expertise in at least one area: model efficiency, real time alignment or algorithmic optimization
Systems thinking ability to understand and optimize across the full ML stack
Strong programming skills in Python. Experience with deep learning frameworks (PyTorch, JAX, TensorFlow)
Knowledge of model optimization techniques (RLHF, finetuning)
Experience in an industry lab with computing at scale
Above all, we're looking for great teammates who make work feel lighter and aren't afraid to go out on a limb with bold ideas. You don't need to be perfect, but you do need to be adaptable. We encourage you to apply, even if you don't check every box.
Flexible work: In-person collaboration in the Bay Area, a distributed global-first team, and quarterly offsites.
Adaption Passport: Annual travel stipend to explore a country you've never visited. We're building intelligence that evolves alongside you, so we encourage you to keep expanding your horizons.
Lunch Stipend: Weekly meal allowance for take-out or grocery delivery.
Well-Being: Comprehensive medical benefits and generous paid time off.
adaption