Associate Director, Applied AI/ML
As Associate Director of Applied AI/ML at Fitch Ratings, I architect and deliver end-to-end AI/ML solutions across Fitch's credit ratings ecosystem. My work focuses on building agentic AI systems, RAG pipelines, LLMOps infrastructure, and production ML solutions that enable analysts to interact with ratings data, financial filings, and regulatory documents through natural language. I lead a 15-person applied AI/ML team, guiding initiatives from discovery through production.
Designed multi-agent workflows and autonomous research agents for credit ratings analysis
Built document intelligence pipelines for large-scale ratings corpora and regulatory documents
Integrated AWS Bedrock, Azure AI Foundry, and OpenAI for scalable LLM serving
Built end-to-end infrastructure for model training, evaluation, deployment, and monitoring
Design and build multi-agent workflows and autonomous research agents using LangChain, LangGraph, and LlamaIndex. These systems enable analysts to interact with ratings data, financial filings, and regulatory documents through natural language, with agents reasoning, retrieving, and synthesizing information autonomously.
Develop document intelligence pipelines using pymupdf4llm for high-fidelity PDF extraction, embedding models for semantic representation, and Azure AI Search for vector retrieval over large-scale ratings corpora.
Integrate AWS Bedrock, Azure AI Foundry, and OpenAI API for foundation model access, prompt management, and scalable LLM serving across multiple internal products. Deploy across hybrid cloud infrastructure: AWS (ECS, Lambda, S3, SageMaker), Azure AI Foundry, and bare metal (Hetzner, Digital Ocean).
Build and maintain full MLOps and LLMOps infrastructure, orchestrating model training, evaluation, deployment, and monitoring with SageMaker, Airflow, Docker, and Kubernetes. Implement distributed task processing and caching layers using Redis and Celery for async ML inference and high-throughput API serving.
Architect and deliver end-to-end AI/ML solutions across Fitch's credit ratings ecosystem
Design multi-agent workflows and autonomous research agents using LangChain, LangGraph, and LlamaIndex
Develop document intelligence pipelines using pymupdf4llm and Azure AI Search for vector retrieval
Build and maintain full MLOps and LLMOps infrastructure with SageMaker, Airflow, Docker, and Kubernetes
Collaborate with business stakeholders, analysts, and technology leadership on high-impact AI/ML opportunities
Mentor junior and senior engineers on AI/ML engineering best practices across a 15-person applied AI/ML team
My work at Fitch Ratings directly enables analysts to leverage AI for credit ratings analysis, transforming how the organization processes and interprets financial data across global capital markets.
Working at the intersection of AI and financial services has deepened my expertise in agentic AI systems, RAG architectures, and enterprise-scale LLMOps. Leading a 15-person team has strengthened my abilities in technical leadership, stakeholder management, and delivering full-stack AI/ML products from prototype to production across hybrid cloud infrastructure.