JPMorganChase AI Research is a global team of research scientists, engineers and product managers that develops novel AI capabilities and partners across the firm to translate breakthrough AI/ML techniques into deployed solutions. The team brings together researchers across multi-agent systems, foundation model training, reinforcement and continual learning, multimodal document AI, formal reasoning, planning and verification, and synthetic data generation - alongside experts in trustworthy AI, privacy, and cryptography - to solve hard problems end-to-end in ways that are secure, reliable, and scalable in financial services.

AI Research sits within the Chief Data & Analytics Office (CDAO) at JPMorganChase which is responsible for accelerating the firm’s data and analytics journey. This includes ensuring the quality, integrity, and security of the firm's data, as well as leveraging this data to generate insights and drive decision-making. The CDAO is also responsible for developing and implementing solutions that support the firm’s commercial goals by harnessing artificial intelligence and machine learning technologies to develop new products, improve productivity, and enhance risk management, effectively and responsibly.

 As a Research Scientist in AI Research, you will work on developing novel techniques, tools, and frameworks to model and solve complex large-scale problems in the bank. You will develop a close understanding of the challenges in practical applications of AI and leverage that understanding to formulate new research directions and solutions that can shift the frontier. Your work will span from early-stage innovation and rigorous evaluation to production-scale delivery in close collaboration with engineering and product teams.

Job responsibilities

• Work on commercially oriented research projects in collaboration with internal data scientists, applied engineering teams and stakeholders across businesses e.g. Commercial & Investment Banking (including Markets), Asset & Wealth Management, Consumer & Community Banking, etc.
• Formulate problems, generate hypotheses, develop new algorithms and models, conduct experiments and rigorous evaluations, synthesize and communicate results, and deliver well-tested, high-quality code.
• Contribute to high-impact business applications, reusable assets and products, and research initiatives.
• Provide thought leadership on internal and external forums through white papers, publications and presentations.

Required qualifications, capabilities, and skills

• Currently pursuing a PhD in Computer science, Engineering, or related fields with relevant research experience in AI/ML
• Research publications in top-tier AI/ML venues (e.g., conferences, journals) – broad conferences such as NeurIPS, ICML, ICLR, etc., or highly regarded specialized conferences such as ICAPS, CRYPTO, etc.
• Ability to deliver modular, optimized, high-quality Python code with tests
• Deep understanding of fundamental AI/ML techniques and a strong grasp of current state of the art in specific areas of expertise
• Practical experience with statistical data analysis, experimental design and creation of meaningful benchmarks and metrics for evaluation.
• Effective verbal and written communication skills with the ability to address both technical and business audiences.
• Motivated to make an impact on problems in the financial services domain.
• Curiosity, creativity, resourcefulness, and a collaborative spirit
• Strong track record of research in one of the following areas: (a) Synthetic Data Generation across multiple modalities (tabular, images, time-series, etc.), synthetic personas and agent simulations (b) Hybrid agent architectures including LLMs and AI planning, formal logic, solvers and neuro-symbolic approaches (c) Multimodal information discovery, deep research and analysis spanning multiple modalities e.g. websites, news, documents, structured data, images, audio, etc. (d) Foundation model training for structured data modalities e.g. tabular, graph, time-series (e) Reinforcement Learning, continual learning for agents, model post-training and test-time adaptation. (f) Advanced multi-agent systems including co-ordination, negotiation and mechanism design. (g) Cryptography and secure multi-party computation with applications to privacy-preserving ML, security and privacy in agentic systems, etc.

Preferred qualifications, capabilities, and skills

• (Depending on specialization) Practical experience with ML libraries (e.g. PyTorch, TensorFlow/Keras, HuggingFace Transformers etc.) and agent frameworks (e.g. LangGraph, Google ADK, etc.), familiarity with common formal verification tools and languages (e.g., Coq/Isabelle/Lean, TLA+, Alloy, Z3, PDDL) 
• Practical software development experience in collaborative project settings such as open-source projects or industry experience.
• Proven ability to translate business requirements into technical problem formations and deliver novel solutions into production.

#AIR_jobs