AI Research Scientist Jobs

View detail

You will work across the model development loop, from research questions to training runs to evaluation. This includes designing and testing architecture changes and training regimes for large language models, running controlled experiments at scale and isolating causal effects, studying failure modes in reasoning, generalisation, robustness, and representation, shaping objectives, data mixtures, and optimisation choices that influence model behaviour, building and refining evaluations that measure capability and reliability, analysing training dynamics using logs, metrics, and model outputs, collaborating with ML systems engineers on distributed training and training operations, and writing clear internal notes that turn experimental results into design decisions. You will spend substantial time in code, training runs, logs, and evaluation outputs with the goal of clarity about what improves the model and why. You will work hands-on with code as a primary tool for thinking, moving between theory and implementation quickly and precisely, preferring controlled experiments over broad sweeps, using logs, metrics, and model behaviour to guide decisions, and working closely with engineering counterparts to scale and validate ideas.

View detail

The role involves working as a research engineer in an AI-related company, being enthusiastic and motivated, collaborating within a team to solve challenging problems, learning and teaching within the team, and operating in a hybrid working culture based on trust.

View detail

The role involves collaborating with Modal’s SDK team and other product engineers to build a framework and process for evaluating agent productivity. Responsibilities include defining quantitative objectives, designing systems to measure performance, translating results into product improvements, staying current with new developments in tools and workflows, and working with customers to understand their use of coding agents with Modal and identify areas for providing more value.

View detail

Design and conduct evaluations of Abridge models and products; engage with external researchers and other stakeholders on designing and conducting research on ambient AI and research that leverages Abridge data; develop a user-centric and patient-centric mindset grounding research in empathy for providers and patients; collaborate with cross-functional product teams to ensure research is informed by current practices and product roadmap; write technical reports and give presentations to internal and external stakeholders; actively contribute to the wider research community by publishing original research in leading peer-reviewed venues; mentor research interns.

View detail

As a senior AI Researcher for reinforcement learning, you will shape and improve the underlying reinforcement learning methodology, maintain a high-quality training codebase, and conduct large-scale experiments to improve performance benchmarks. Your responsibilities include conducting large-scale LLM training runs, analyzing evaluation scores, proposing and implementing improvements, staying at the forefront of reinforcement learning research by identifying and iterating on novel approaches, optimizing RL training loops to scale training infrastructure, and collaborating cross-functionally with other post-training teams to convert feedback into actionable training signals for measurable improvements in performance.

View detail

Conduct original research to advance the state of the art in machine learning and artificial intelligence. Design, implement, and evaluate novel algorithms, models, or training approaches at large scale. Collaborate with researchers and engineers to translate research insights into production systems and real-world applications.

View detail

The ML Research Scientist will use AI and Machine Learning to transform sensor data into personalized intelligent health and fitness experiences by working closely with a cross-functional R&D and production team to prototype and ship solutions. Projects include advancing the Pod’s adaptive thermoregulation system using reinforcement learning and closed-loop control, developing multimodal health foundation models integrating physiology and environmental context from Pod signals, wearable sensors, and contextual data, and building high-fidelity physiological simulators to model how daily behaviors affect sleep and readiness. The scientist will tackle problems with a systems approach and make data-driven decisions to deliver the best products to users.

View detail

This role leads the Automated Red Teaming (ART) effort by building scalable, research-driven systems that continuously discover failure modes in models and mitigations, translating findings into actionable, production-facing improvements to reduce expected harm by identifying high-leverage weaknesses early and reliably. The responsibilities include owning the research and technical direction for automated red teaming across catastrophic risk areas initially focused on automated classifier jailbreak discovery (cyber and bio), automated bio threat-development elicitation, and Chain-of-Thought monitoring evasion probing. The role requires tight partnership with vertical risk teams to define threat models, prioritize targets, and implement mitigations; collaboration with the Classifiers team to convert discovered attacks into training data, evaluations, and robustness improvements; and working with product, engineering, and safety stakeholders to ensure outputs are operationally useful.

View detail

The Research Engineer will help design, evaluate, and productionize next generation AI inference systems by working at the intersection of applied research and real-world deployment to develop techniques that improve performance, efficiency, reliability, and cost of large-scale inference workloads. They will collaborate with systems engineers, ML engineers, and infrastructure teams to translate research ideas into practical implementations, writing and optimizing performance-critical kernels and improving low-level execution efficiency on modern accelerators. Responsibilities include researching, implementing, and evaluating state-of-the-art techniques for AI inference such as speculative decoding, prefill–decode disaggregation, quantization, and kernel-level optimizations focused on real-world customer use cases; designing and running experiments to understand trade-offs across latency, throughput, cost, and quality; analyzing real-world inference workloads to identify opportunities for improvements; staying current with advances in ML systems and inference research; sharing findings through internal reports and external contributions; and defining and contributing to the company roadmap to impact products and customers.

View detail

As a Researcher for cybersecurity risks, you will design and implement mitigation components for model-enabled cybersecurity misuse, including prevention, monitoring, detection, and enforcement, under guidance from senior technical and risk leadership. You will integrate safeguards across product surfaces in partnership with product and engineering teams to ensure protections are consistent, low-latency, and scalable with usage and new model capabilities. You will evaluate technical trade-offs within the cybersecurity risk domain such as coverage, latency, model utility, and user privacy, proposing pragmatic and testable solutions. You will collaborate closely with risk and threat modeling partners to align mitigation design with anticipated attacker behaviors and high-impact misuse scenarios. You will execute rigorous testing and red-teaming workflows, stress-testing the mitigation stack against evolving threats including novel exploits, tool-use chains, and automated attack workflows, iterating based on findings.