Red-team frontier benchmarks: identify and reproduce issues like reward hacking, grader failures, task ambiguity, shortcuts, and environment vulnerabilities.
Develop benchmark repairs: propose next steps, implement fixes with authors, and verify without breaking valid solutions.
Drive research impact: contribute to technical reports, open-source tools, publications, and new approaches to benchmark evaluation.
Snorkel AI is the frontier AI data lab, helping teams build the data and environments behind high-performing frontier and agentic AI. Founded out of the Stanford AI Lab in 2019, Snorkel works with leading AI labs and enterprises and is scaling rapidly.
Build and post-train agents for scientific research, including literature search, code execution, and domain tool use.
Develop RL and multi-agent methods for proposing, testing, and critiquing hypotheses, with strong baselines and evaluation.
Design verifiers, reward signals, and reproducible experimental loops for scientific workflows, ensuring evidence-based improvements.
Prime Intellect builds the open superintelligence stack, providing infrastructure for frontier AI labs to train, evaluate, and deploy models at scale. The company has raised $150M from top investors and operators, and is building a culture focused on frontier research, real infrastructure, and go-to-market for a new category.
Act as the subject-matter specialist for AI-enabled Software Development and Software Quality Assurance.
Research and benchmark AI applications, tools, and industry practices.
Provide expert advice on practical AI applications across the Software Development Lifecycle.
XM is a global financial services and trading company. With engineering teams across multiple locations, the company focuses on software development and quality assurance in a fast-paced environment.
Own the fellowship curriculum and learning design, ensuring alignment with AI industry practices and market demand.
Oversee the quality of all learning materials and improve continuously using learner outcomes and feedback.
Build and maintain the instructor network by sourcing, interviewing, and calibrating instructors to uphold program standards.
Anyone AI runs an AI fellowship program providing practical, industry-aligned education. The company values a dynamic, fast-iterating culture and maintains rigorous standards for teaching and learning quality.
Design and build challenging evaluations and executable environments for coding and long-horizon agents.
Develop scoring logic and verifiers to detect reward hacks and trace failures through model outputs and harness code.
Build and scale RL data pipelines for generating and curating challenging coding and agentic tasks.
Prime Intellect is building the open superintelligence stack, providing infrastructure for post-training frontier AI models and open-source models trained for long-horizon tasks. The company has raised $150M from top investors and is a mission-driven team passionate about accelerating science and AI.
Design evaluations of conceptual reasoning in AI safety, covering governance, alignment, macrostrategy, and economics.
Develop rubrics that separate genuine reasoning from plausible-sounding filler.
Iterate on tasks with frontier models and collaborate with a top research team.
Trajectory Labs is a startup automating AI safety research. We build safety and alignment evals, red-teaming programs, and RL environments for frontier AI labs. As a one-year-old company, we've been featured in major frontier lab safety reports and led high-impact red-teaming efforts.
Build evaluation frameworks for AI agents, including datasets and metrics for real-world task performance.
Design experiments to compare models and tools, identifying regressions and trade-offs.
Investigate agent failures and improve performance across models, harnesses, and tooling.
Gauntlet builds the financial systems of the future, operating across the onchain finance stack to offer vault products. They serve over $1.7B in client TVL across fintechs, protocols, exchanges, and capital allocators, combining traditional finance and crypto-native expertise.
Apply formal methods to measure and improve policy enforcement for AI agents.
Build automated checks for policy changes using simulations and solver-backed analysis.
Ship research results into the product quarterly, with metrics and rollback capabilities.
Sondera develops AI agents that enforce natural-language policies. The company is backed by Decibel, Runtime Ventures, and Aviso Ventures, and its team has built products at Cylance and Databricks.
Contribute to post-training and evaluation infrastructure for a novel multimodal AI system.
Fine-tune models using supervised and reinforcement learning approaches (e.g., SFT, RLAIF, RLVR).
Define custom metrics and develop tools to assess and mitigate AI-related risks in generative AI.
Sony AI is Sony's research organization using AI to unleash human creativity, working with gaming, entertainment, and music divisions. The Ethical Creative and Protective AI Lab operates as a globally distributed team across the U.S. and Canada, with some 900 million Sony devices in use worldwide.
Design pretraining, SFT, and RL data for engineering tasks across CAD, CAE, and EDA.
Define evaluation criteria and agent architectures for multi-step engineering workflows.
Build benchmarks and diagnostics to identify model failures and trace back to gaps in data or agent design.
Mistral is a full-stack AI company building frontier models, developer tools, and compute for enterprises. We are a diverse, dynamic team distributed globally, passionate about AI and committed to low-ego collaboration and innovation.
Own Open Benchmarks Grants end-to-end, leading operations for Snorkel’s $30M commitment to open benchmarks.
Lead external research partnerships and manage budgets, resources, and cross-functional execution.
Build and scale a multi-year research program, measuring impact to improve future cycles.
Snorkel AI is a frontier AI data lab that helps teams build datasets, benchmarks, and evals for high-performing AI systems. Founded out of Stanford AI Lab in 2019, the company works with leading AI labs and enterprises, offering a fast-paced, collaborative high-growth culture.
Design and build AI-powered tools to decode complex biology and accelerate pharmaceutical research.
Develop reliable, scalable systems with LLMs, agents, and data pipelines in a fast-paced environment.
Collaborate with scientists and engineers to give biomedical researchers intuitive, powerful tools.
Owkin is an agentic AI company pioneering Biological Artificial Superintelligence to solve problems in biology. It offers a friendly and informal working environment with an international team of high technical and scientific backgrounds.
Design, build, and test Claude skills for IT and Security workflows with connected MCP servers.
Assist in scoping, building, and iterating on AI agents to automate repetitive IT and Security tasks.
Support security operations, IT helpdesk, and documentation maintenance in Confluence.
Azul is a technology company that provides Java platforms and cloud infrastructure solutions. It is a global organization focused on innovation and offers a collaborative culture with opportunities for professional growth.
Lead end-to-end planning and delivery of Agog's labs, including partner coordination, participant selection, and logistics.
Manage relationships with residency and fellowship hosts, ensuring alignment with Agog's mission and program goals.
Develop and manage a pipeline of nominees for immersive educational opportunities and cultivate an engaged alumni network.
Agog is an immersive media institute helping creators and nonprofit leaders use extended reality (XR) technologies to open new avenues for empathy, understanding, and activism. Founded by Wendy Schmidt and Chip Giller with an accomplished advisory board, Agog is poised for big impact as an ecosystem builder across the XR industry, nonprofit sector, philanthropy, and academia.
Design, build, and maintain certification programs for system integrator partners across infrastructure and agentic tracks.
Develop deep subject matter expertise in deploying AI and LLM workflows from prototype to production.
Collaborate with Forward Deployed Engineering and Product teams to ensure certification content stays current and adapts into the core product.
Cohere is a leading security-first enterprise AI company that builds cutting-edge foundation models and end-to-end products for real-world business problems. It is a global team of researchers, engineers, designers, and more, headquartered in Toronto with offices across multiple continents.
Build post-training and evaluation systems for multimodal agents operating browsers and desktops.
Turn real workflows into environments, rewards, and experiments that drive reliable agent improvement.
Work directly with customers and research partners while maintaining a substantive research agenda.
Prime Intellect is building the open superintelligence stack, providing infrastructure for post-training frontier AI models. The company has raised $150M in funding and fosters a culture of ownership and collaboration for ambitious AI teams.
Own quality for non-deterministic, multi-agent AI systems, building the evaluation harness that underpins AI governance.
Verify grounding and faithfulness, building automated checks to stop hallucinations from reaching production.
Red-team for prompt injection, jailbreaks, and tool misuse, and prove multi-tenant data isolation.
SOFICO Intelligence uses AI to extract value from back-office systems, building agentic AI for financial, leasing, fleet, and mobility data. Backed by global ERP leader SOFICO, it's a remote-first team of curious, self-directed builders.
Own applied RL programs from ambiguous customer problem to deployed improvement loop.
Understand workflows, build environments and evaluations, run post-training experiments, and deploy results.
Work directly with customer researchers, engineers, and leaders, accountable for research result and practical impact.
Prime Intellect builds the open superintelligence stack, providing a full-stack platform for frontier-scale post-training of AI models. The company has raised $150M from top investors and is seeking builders who want to work at the intersection of frontier research, real infrastructure, and go-to-market.
Own the path from emerging research artifact to excellent open-source and product experience.
Build core open-source systems and shape researcher-facing features in our training platform.
Work with external researchers and contributors to turn fast-moving abstractions into stable public interfaces.
Prime Intellect is building the open superintelligence stack, providing infrastructure and a full-stack platform for post-training frontier AI models. The company has raised $150M in funding from prominent investors and aims to make open-source AI research fully reproducible and accessible.
Research new attack surfaces and threat vectors, including AI coding agents, agent plugins, and package ecosystems.
Build AI-powered workflows and agents that automate research, analysis, and detection.
Analyze software supply chain incidents and events and write up findings as blog posts and conference presentations.
GitLab is the intelligent orchestration platform for DevSecOps. More than 50 million registered users and more than 50% of the Fortune 100 trust GitLab to ship better, more secure software faster.