Own a multimodal ML work-stream from problem definition through deployment, translating clinical needs into clear ML objectives and building systems with transformers, self-supervised learning, and segmentation.
Design rigorous evaluations beyond offline metrics, partner with engineering to productionize models, and investigate failure modes like laterality errors and hallucination.
Communicate research findings clearly, mentor less experienced researchers, and contribute to the research roadmap by identifying promising approaches.
Design, develop, train, and evaluate machine learning models for interpreting complete CT studies.
Research foundation-model approaches for medical imaging, including 3D and volumetric learning at large scale.
Develop vision-language models connecting medical images with radiologist terminology and reporting patterns.
Our partner is a research-driven team building next-generation AI models to understand complete CT studies. They work with a dataset of approximately 10 million patients and collaborate with fellowship-trained radiologists across multiple specialties.
Generative media synthesis: building the models behind features like Video Regenerate, lipsync, and video translation — photorealistic video and audio synthesis as production research, not demos.
Zero-shot voice and roomtone cloning: realistic cloning from only a few minutes of reference audio.
Multimodal understanding: vision-language systems that let Descript's agentic editing features reason over visual and audio content, plus the evals that balance quality against cost and latency.
Descript is building a simple, intuitive, fully-powered editing tool for video and audio — an editing tool built for the age of AI. They are a team of 150 with backing from top investors, and each new employee has a measurable influence on the company's direction.
Own core computer vision models from 3D human pose to body modeling, shipping them to production.
Manage the full data lifecycle including training frameworks, annotation workflows, and evaluation pipelines.
Extend movement understanding into multimodal AI and automate model training and deployment systems.
Sword is building AI to heal billions, pioneering an AI-native care platform that delivers personalized, connected healthcare without traditional barriers. Backed by $500M and with over 1 million members and 2,000+ enterprise clients, Sword is a fast-growing frontier AI lab with a culture of clinical rigor and AI fluency.
Conduct cutting-edge machine learning research, building and training large language models.
Focus on research projects aimed at expanding the frontier of knowledge in language modelling and associated areas.
Disseminate your research results through publications, datasets, and code.
Cohere is a security-first enterprise AI company that builds cutting-edge foundation AI models and end-to-end products for enterprise AI systems. We are a global team of researchers, engineers, and designers passionate about our craft, with offices in Toronto, San Francisco, London, and more.
Work at the forefront of AI-powered drug discovery, advancing frontier research and end-to-end applications across discovery, design, and translation.
Collaborate with a cross-functional team of top-notch scientists, engineers, and drug hunters in the quest of scalable biomedical discovery.
Leverage an unparalleled array of infrastructures including one of the largest AI compute clusters in biopharma and industrial-scale phenomics.
Recursion is a clinical-stage TechBio company decoding biology to radically improve lives. They operate one of the largest AI compute clusters in biopharma and offer a vibrant, cross-functional environment with frontier research and real-world delivery.
Design and develop assistive AI technology for interactive 3D workflows.
Build NLP pipelines and knowledge graphs to extract intent and structure from user input.
Integrate cloud and client LLMs to drive creative workflows and ship features end-to-end.
Intangible is a spatial intelligence company for creatives in industries where space and movement matter. It is a venture-funded, remote-first interdisciplinary team of engineers, artists, and designers who have built products at major tech and entertainment companies.
Evaluate LLM architecture logic for technical accuracy and audit ML code and notebooks for efficiency.
Refine RLHF frameworks to align models with human intent and analyze model reasoning in complex chain-of-thought prompts.
Benchmark performance by conducting comparative testing between model outputs based on technical metrics.
Prolific connects researchers with a global pool of participants for collecting high-quality human data to train AI models. With over 35,000 users, they focus on ethical data gathering to advance AI capabilities.
Design, develop, and optimize computer vision models and deep learning capabilities across the product portfolio.
Build and maintain ETL pipelines, clean datasets, and perform feature engineering for machine learning applications.
Collaborate with cross-functional teams to productionize, deploy, and monitor models in real-world environments.
FloVision is a remote-first startup improving the food supply chain with computer vision and ML. It is a Series A company with a distributed team across the US, UK, and Ireland.
Design and build production-grade AI agents and LLM-powered systems.
Develop multimodal data ingestion and structured information extraction pipelines for unstructured data.
Own AI solutions end-to-end from requirements to deployment and optimization.
Mutt Data is a dynamic startup committed to crafting innovative systems using cutting-edge Big Data and Machine Learning technologies. They have a remote-first culture with values centered on data nerdiness, teamwork, ownership, and positivity.
Develop and deploy deep learning models to predict therapeutic intervention effects.
Design experiments and collaborate with cross-functional teams including biologists and wet lab scientists.
Validate models and communicate results to diverse audiences to drive informed decision-making.
Absci is a clinical-stage biotechnology company advancing novel therapeutics using generative AI. It is a global company with offices in the US, Switzerland, and Serbia, fostering a collaborative and innovative culture.
Architect end-to-end computer vision pipelines for real-world safety detection, including object detection, tracking, and semantic segmentation.
Drive the technical roadmap for edge and cloud perception, optimizing models for constrained hardware without sacrificing accuracy.
Mentor senior scientists and engineers, translating customer needs into precise engineering challenges.
Samsara is the pioneer of the Connected Operations™ Cloud, a platform that enables organizations dependent on physical operations to harness IoT data for actionable insights and improved safety, efficiency, and sustainability. As a recently public company, they foster a culture of rapid career development, collaboration, and innovation while scaling globally.
Build the software backbone for autonomous foundation models, including multimodal data pipelines and training workflows.
Implement and iterate on LLM, VLM, and VLA architectures, owning model code paths and inference runners.
Deliver production-grade serving tooling for low-latency operation and own systems from architecture to iteration.
We're building the next generation of ground transportation with advanced physical AI to simplify freight challenges. Our stealth team, founded by engineers who scaled autonomous driving, is developing a new vehicle platform and focuses on creating reliable real-time autonomous systems.
Own ML projects end-to-end from exploration to production deployment.
Build agentic LLM systems with multi-step workflows and tool use.
Treat evaluation as core engineering work with robust testing pipelines.
We are building AI to heal billions and unlock humanity's full potential, pioneering a new approach to healthcare. With over 1 million members, 15 million AI sessions, and $500 million raised, we are defining a new standard for healthcare.
Support training, fine-tuning, and evaluation of neural foundation models.
Build and maintain data pipelines for petascale neurobehavioural datasets.
Prototype tools and demos to apply trained models to robotics and embodied AI tasks.
Netholabs builds AI grounded in biological intelligence by recording petascale neurobehavioural data to train neural foundation models. They are a small, fast-moving research team focused on AI, robotics, and personalized intelligence.