Lead Machine Learning Engineer - Agentic Pricing

JPMorganChase; Jersey City (NJ); On-site; USD 171,000 - 260,000; Full time

Job summary

JPMorganChase is building production AI agents on the NEO platform within the Corporate & Investment Bank tech organization. You will design, productionize, and operate LLM-powered agents across CIB and Payments, turning real use cases into shipped systems.

The role focuses on agent workflows, retrieval and memory design, rigorous evaluation, guardrails, and secure deployment on public cloud (AWS/Azure), supporting end-to-end delivery from prototype to production.

Qualifications

• MS in Computer Science, Statistics, Mathematics, Machine Learning, or related field (or equivalent).
• Minimum 7 years of development experience, with at least 4 years on AI/ML solutions.
• Hands-on production experience building LLM-powered or agentic applications, with tracing, evaluations, and guardrails.
• Strong programming skills in Python and knowledge of data structures, algorithms, ML, data mining, IR, and statistics.
• Practical RAG experience, including retrieval quality, embeddings, and vector stores.
• Expert knowledge of AWS, Azure, or Kubernetes.

Responsibilities

• Design and ship production agents on NEO across the federated portfolio, from prototype to production.
• Build retrieval that performs in production via chunking, ranking, and grounding for accurate, auditable answers.
• Design agent memory with episodic and semantic memory as memory nodes, with recall and decay policies.
• Own context management with entitlement-, lineage-, and tenant-aware context for restricted information access.
• Compose multi-agent workflows using A2A, integrating tools and data through MCP servers (Bitbucket, Confluence, Databricks, Kubernetes, Snowflake, Splunk).
• Build and run evaluations, including task-level and end-to-end agent evaluations, regression suites, and quality/safety gating.

Skills

Python

AWS

Azure

Kubernetes

SQL

NoSQL

OpenSearch

Redis

Embeddings

LLM-as-judge

Education

MS in Computer Science or related field

Tools

Databricks

MCP

Bitbucket

Confluence

Snowflake

Postgres

OpenSearch

Redis

Job description

JPMorganChase is building production AI agents through its NEO agent runtime platform within the Corporate & Investment Bank technology organization under Digital & Platform Services / Data Analytics. In this role, you will design, productionize, and operate LLM‑powered agents across the Corporate & Investment Bank and Payments, helping turn real use cases into shipped systems.

NEO already supports a federated portfolio of production agents, with additional capabilities expanding over time. You will focus on agent workflows, retrieval and memory design, rigorous evaluation and guardrails, and secure deployment on public cloud, supporting end‑to‑end delivery from prototype to production.

What you’ll do

• Design and ship production agents on NEO across the federated portfolio, owning work from prototype through production.
• Build retrieval that performs in production through chunking, ranking, and grounding strategies that support accurate and auditable answers.
• Design agent memory using episodic and semantic memory organized as memory nodes, including recall, summarization, and decay policies tuned to each use case.
• Own organizational context management by assembling entitlement-, lineage-, and tenant‑aware context so agents reason over permitted information only.
• Compose multi‑agent workflows using A2A, integrating tools and data through MCP servers (Bitbucket, Confluence, Databricks, Kubernetes, Snowflake, Splunk).
• Build and run evaluations, including task‑level and end‑to‑end agent evaluations, regression suites, LLM‑as‑judge, and quality and safety gating prior to release.
• Deploy and operate solutions on public cloud (AWS and/or Azure) using strong SDLC, security, resiliency, and observability practices.
• Collaborate with product and business partners across the Corporate & Investment Bank and Payments to take use cases from concept to supported agents.

What you bring

• Education: MS in Computer Science, Statistics, Mathematics, Machine Learning, or related field (or equivalent experience).
• Experience: minimum 7 years of development experience, with at least 4 years working on AI/ML solutions.
• Hands‑on production experience building LLM‑powered or agentic applications, including tracing, evaluations, and guardrails.
• Strong programming skills in Python, plus deep knowledge of data structures, algorithms, machine learning, data mining, information retrieval, and statistics.
• Practical RAG experience, including retrieval quality, embeddings, and vector stores.
• Expert knowledge of at least one of: AWS, Azure, or Kubernetes.
• Knowledge of data management and data model design; real‑time processing with both SQL (for example, Postgres) and NoSQL stores (for example, OpenSearch and Redis).
• Excellent communication skills and the ability to partner effectively with senior technical and business stakeholders.

Technology areas you may work with

• NEO, AWS, Azure, Python
• Databricks, GenAI Gateway, MCP
• Bitbucket, Confluence, Kubernetes, Snowflake, Splunk
• SQL, Postgres, NoSQL, OpenSearch, Redis
• A2A, LLM‑as‑judge, vector stores, embeddings
• Graph RAG, knowledge graphs, graph databases
• JavaScript, TypeScript, Next.js, Svelte, AG-UI, NEO UI SDK
• Go, Rust

Team context

• J.P. Morgan’s Commercial & Investment Bank is a global leader across banking, markets, securities services and payments.
• Corporations, governments, and institutions in more than 100 countries entrust their business to the firm.
• The Commercial & Investment Bank provides strategic advice, raises capital, manages risk, and extends liquidity in global markets.

Location: Jersey City, NJ (onsite). Salary: USD 171,000 - 260,000 per year.