Principal Software Engineer - Data and AI - Accelerator Business

JPMorgan Chase JPMorgan Chase · Banking · LONDON, LONDON, United Kingdom · Corporate Sector

Principal Software Engineer focused on designing and developing scalable, self-service solutions for GenAI applications, including RAG pipelines and agents with planning, memory, and orchestration. The role involves implementing tools for model management, monitoring performance and drift, ensuring compliance, deploying AI services to cloud infrastructure, and designing microservices architectures with instrumented agents.

What you'd actually do

  1. Design and develop scalable, self-service solutions for documentation, SDKs, configurations and pipelines to enable rapid deployment of GenAI applications (including Retrieval-Augmented Generation (RAG) pipelines) and agents with planning, memory, and workflow orchestration
  2. Implement tools and frameworks for model versioning, experiment tracking, and lifecycle management
  3. Develop systems to monitor model performance and address data and model drift
  4. Recommend best practices for model integration and deployment patterns
  5. Design and implement effective testing strategies, including unit, component, integration, end-to-end, performance, and champion/challenger tests, establish output validation best practices, recommendations and guardrails to reduce hallucinations.

Skills

Required

  • Java and/or Python programming languages
  • Deployed production systems to GenAI platforms such as Google VertexAI, OpenAI, AWS Bedrock, or LangChain
  • Utilized cloud technologies (AWS/Azure/GCP), distributed systems, CI/CD tools, infrastructure-as-code tools, and containerization/orchestration tools (Docker, Kubernetes) to operate, support, and secure mission-critical applications
  • Previous experience deploying and managing LLM-model based applications and agents
  • Exposure to vector stores such as Pinecone, GCP RAG engine, and AWS S3 Vector Buckets
  • Exposure to cloud-native microservices architecture
  • Familiarity with advanced AI/ML concepts and protocols, including Retrieval-Augmented Generation (RAG), agentic system architectures, and Model Context Protocol (MCP)
  • Hands-on experience with agentic frameworks (LangChain, CrewAI, AutoGen, LangGraph, ADK)
  • Strong communication skills for both technical and non-technical audiences.

Nice to have

  • Experience working in highly regulated environments or industries
  • Experience with distributed computing, data sharding, and performance optimization.
  • Demonstrated experience in financial services, particularly retail banking operations.

What the JD emphasized

  • Ensure platform compliance with data privacy, security, and regulatory standards

Other signals

  • GenAI applications
  • Retrieval-Augmented Generation (RAG) pipelines
  • agents with planning, memory, and workflow orchestration
  • model versioning
  • experiment tracking
  • lifecycle management
  • monitor model performance
  • data and model drift
  • model integration and deployment patterns
  • testing strategies
  • output validation best practices
  • guardrails to reduce hallucinations
  • platform compliance
  • data privacy
  • security
  • regulatory standards
  • scalable AI services to cloud infrastructure
  • monitoring
  • observability for agent performance
  • microservices-based architectures
  • orchestrate multi-step workflows
  • instrument agents for tracing
  • metrics
  • feedback loops
  • reliability and utility
  • Java and/or Python programming languages
  • production systems to GenAI platforms
  • Google VertexAI
  • OpenAI
  • AWS Bedrock
  • LangChain
  • cloud technologies (AWS/Azure/GCP)
  • distributed systems
  • CI/CD tools
  • infrastructure-as-code tools
  • containerization/orchestration tools (Docker, Kubernetes)
  • mission-critical applications
  • LLM-model based applications and agents
  • vector stores
  • Pinecone
  • GCP RAG engine
  • AWS S3 Vector Buckets
  • cloud-native microservices architecture
  • advanced AI/ML concepts and protocols
  • Retrieval-Augmented Generation (RAG)
  • agentic system architectures
  • Model Context Protocol (MCP)
  • agentic frameworks (LangChain, CrewAI, AutoGen, LangGraph, ADK)