Lead Generative AI Developer

About the Role

We are looking for a Lead Generative AI Developer to join our COO Technology Division in New York. In this high-impact role, you will architect, develop, and operationalize cutting-edge Generative AI and Large Language Model (LLM) solutions that directly transform how Citi's operational teams work. You will collaborate with cross-functional stakeholders - including operations leads, data engineers, product managers, and enterprise architects to deliver enterprise-grade AI capabilities at scale.

This is a hands-on engineering role for a builder who thrives at the intersection of applied AI research and production software engineering.

Key Responsibilities

• Design & Build GenAI Solutions: Architect and implement end-to-end Generative AI pipelines including LLM integrations, Retrieval-Augmented Generation (RAG) systems, autonomous AI agents, and prompt engineering frameworks.

• Python Development: Develop robust, scalable, and production-ready Python services and APIs that power AI-driven features across COO platforms.

• Model Integration & Fine-tuning: Evaluate, integrate, and fine-tune LLMs (e.g., GPT-5, Claude, Mistral) and embedding models for domain-specific financial use cases.

• MLOps & Deployment: Build and maintain ML/GenAI deployment pipelines using modern MLOps practices, ensuring reliability, observability, and governance.

• Agentic Workflows: Design and implement multi-agent orchestration frameworks (e.g., LangGraph, Google ADK) for complex, multi-step operational workflows.

• Enterprise AI Governance: Collaborate with Citi's AI Risk and Compliance teams to ensure all AI solutions align with regulatory requirements, responsible AI frameworks, and data privacy standards.

• Data Engineering: Design and optimize data pipelines feeding AI systems, working with vector databases (e.g., Pinecone, Weaviate, pgvector) and enterprise data platforms.

• Technical Leadership: Mentor junior developers, lead code reviews, and contribute to GenAI standards and best practices across the COO Technology organization.

• Stakeholder Collaboration: Translate complex business requirements from COO operations stakeholders into technical AI solutions, providing clear communication of trade-offs and timelines.

Required Qualifications

• Experience: 10+ years of professional software engineering experience, with at least 2+ years focused on Generative AI / LLM application development.

• Python: Expert-level Python proficiency — including async programming, API development (FastAPI, Flask), and software design patterns.

• GenAI & LLM Stack:

  • Deep hands-on experience with LLM frameworks: LangChain, LangGraph, LlamaIndex etc

  • Hands on experience with Google Cloud AI Platform

  • Proven experience with RAG architectures, embedding pipelines, and vector search

  • Strong understanding of prompt engineering, few-shot learning, and chain-of-thought techniques

  • Experience integrating with LLM APIs: OpenAI, Azure OpenAI, Anthropic, AWS Bedrock, or Google Vertex AI

• Machine Learning: Solid grounding in ML fundamentals; familiarity with model evaluation, fine-tuning (LoRA, PEFT), and inference optimization.

• Cloud Platforms: Hands-on experience with at least one major cloud provider — AWS, Azure, or GCP — particularly managed AI/ML services.

• Data & Databases: Proficiency with SQL, NoSQL, and vector databases (Pinecone, Weaviate, Chroma, pgvector).

• Software Engineering Practices: Strong understanding of CI/CD pipelines, containerization (Docker, Kubernetes), version control (Git), and automated testing.

• Financial Services Acumen (Preferred): Prior experience in banking, fintech, or a regulated industry is a strong plus.

Preferred Qualifications

• Experience with multi-agent orchestration frameworks (MS AgentFramework, ADK, Strands, LangGraph)

• Familiarity with MLflow, Weights & Biases, or similar experiment tracking and model management tools

• Knowledge of responsible AI practices: bias detection, explainability, hallucination mitigation

• Exposure to Kafka, Spark, or Airflow for data pipeline engineering

• Experience working in an Agile/SAFe delivery environment

• Advanced degree (M.S.) in Computer Science, AI/ML, or a related discipline — or equivalent demonstrated experience

T