Top AI Research Jobs in NYC, NY

Supports AI safety, robotics, cybersecurity, Physical AI, and advanced mobility projects. Conducts research on standards, regulations, and emerging technologies; assists with safety assurance and consulting activities; develops reports, presentations, workshops, and client deliverables; and contributes to AI safety, cybersecurity, robotics, and technology capabilities.

Manage and coach a team of mentors to ensure delivery outcomes for ~200 AI fellows at scale. Build delivery systems, run operational cadence, lead live instruction when needed, translate field signals into curriculum changes, and partner with strategic partners to close the feedback loop between mentors, fellows, and the curriculum team.

Research and develop neural 3D reconstruction models that recover geometry, pose, deformability, and physical properties of manipulated objects from monocular egocentric video. Build generative refinement, contact-aware reconstruction, and simulatable-object prototypes using deep learning, 3D geometry, and differentiable rendering. Define research direction, evaluate challenging real-world data, prototype in Python and PyTorch, and hand successful methods to an integrations researcher for productionization.

Conduct cutting-edge machine learning and AI research applied to Pinterest products. Interns lead projects, develop and evaluate production ML systems, work across algorithms, data engineering, infrastructure, training, inference, and product integration, and write sustainable code. They will use AI-assisted development tools, validate generated outputs for correctness and security, investigate bugs, collaborate with researchers and engineers, and communicate solutions to open-ended technical problems.

Track and evaluate emerging LLM and AI research, assess models and techniques, build prototypes, and collaborate with engineers to deploy useful methods in client systems. Advise product direction, establish metrics for accuracy, reliability, and cost, and communicate research findings clearly. The role requires strong applied research, engineering, judgment, and communication skills in a fast-changing early-stage environment.

Develops reinforcement learning algorithms, model architectures, and experimental infrastructure for frontier-scale training. Responsibilities include scaling RL across model size, context length, and compute; diagnosing numerical, algorithmic, and systemic failures; optimizing performance and cost; and managing end-to-end large-scale runs. The role requires rigorous experimentation, distributed training expertise, quantitative modeling, and strong programming across the ML systems stack.

Leads technical foresight on how AI and emerging technologies may reshape global finance. Assesses AI capabilities, architectures, scaling, constraints, governance, failure modes, and systemic risks; translates findings into scenarios, hypotheses, briefings, research priorities, and thought leadership. Partners with research, standards, advocacy, and industry stakeholders to stress-test assumptions and evaluate technology’s implications for investment workflows, institutions, markets, professional roles, and standards.

Join a research and advisory pipeline for an asset-management startup building an auditable reasoning engine. Advise on agentic AI, world models, interpretability, and systemic risk. Help shape research direction, identify failure modes and guardrails, and provide scholarly evidence (papers, GitHub, Google Scholar) and a short written response to demonstrate expertise.

Lead applied research projects that move LLM-based code models from research to product: build and train code models, create datasets from public and internal code, run large-scale ML experiments, collaborate with product and engineering teams, and evaluate and iterate AI for developer tools.

Build and maintain clinical AI benchmarks, evaluation infrastructure, grading systems, and production harnesses. Convert consented patient cases into de-identified evaluation and training data through clinician review workflows. Lead clinical evaluation studies from protocol design through publication, collaborate on reward design and data curation, and develop agentic evaluations covering tool use, long-horizon tasks, memory, records, labs, and escalation behavior.