Aniruddha Choudhury He
Lead Ai Engineer @American Express
Signup · Get unlimited contacts
WORK HISTORY
Lead Ai Engineer @American Express
Bengaluru, IN
Leading a Pod of 5 AI Engineers with cross collaboration with Business to architect and deploy the Agentic SRE Playbook, automating the lifecycle of Bridge Incidents for the GI. Architected a Multi-Agent \"Triage Swarm\" using LangGraph to autonomously perform Upstream/Downstream impact analysis across hybrid Kubernetes clusters during P1/P0 payment outages. Developed a Federated MCP (Model Context Protocol) Platform to grant LLM agents secure, read-only access to legacy DW data, transaction logs, and Elastic observability traces. Authored Architectural Decision Records (ADRs) for Agent Isolation, ensuring that auto-remediation agents operate within strictly defined \"blast radius\" to prevent cascading failures in production. Built \"Dependency-Aware\" Agents capable of mapping real-time Kubernetes service meshes to identify if a CCP latency spike is rooted in a downstream NoSQL database or an upstream API Gateway. Engineered Agent-to-Agent (A2A) Handshakes where a \"Monitoring Agent\" negotiates with a \"Network Agent\" to verify BGP routing issues before escalating to a human-led Bridge call. Deployed Federated Search with Copilot Studio, consuming MCP connectors to provide SREs with a unified \"Incident Context\" dashboard across Graph relationships (Merchant-to-Terminal) and NoSQL logs. Implemented \"Human-in-the-Loop\" LangGraph Nodes for high-risk write actions (e.g, K8s pod restarts or traffic draining), ensuring AI-driven remediation is governed by senior SRE oversight.9.Architected A2A (Agent-to-Agent) negotiation. Engineered a secure A2A communication protocol. Autonomous agents negotiate resource allocation and incident hand-offs via encrypted gRPC and JSON-Schema.Federated MCP Governance: Architect the Federated Model Context Protocol (MCP) platform. You define how agents discover and securely connect to on-prem DB and cloud (Postgres/Elastic) data without exposing direct database credentials.
EDUCATION
University of Michigan
Mathematics and Computer Science
KIIT - Kalinga Institute of Industrial Technology
Bachelor's degree, Mechanical Engineering
Indian School of Business
Leadership with AI, Computer and Information Sciences and Support Services
ABOUT ANIRUDDHA CHOUDHURY HE
I’m Aniruddha Choudhury, an AI/ML architect, book author, and research-driven innovator with over 10 years of experience building cutting-edge, production-grade machine learning systems that sit at the intersection of Generative AI, MLOps, LLMOps, NLP, Computer Vision, and real-world business outcomes.My strength lies in designing end-to-end AI ecosystems frameworks to multi-agent orchestration systems and domain-specific AI copilots. I specialize in translating foundational research into robust GenAI and Agentic RAG applications that go beyond chatbots—focusing on reasoning, automation, knowledge integration, and long-term scalability.With extensive hands-on experience across NLP, deep learning, computer vision, I have implemented intelligent solutions in diverse domains, and LLM-powered enterprise copilots. I’ve deployed these systems with a strong architectural foundation in MLOps and LLMOps across AWS, Azure, & GCP, ensuring reproducibility, reliability, and rapid iteration.Where I differentiate is not just in delivery—but in thought leadership and innovation. I’m a firm believer that the future of AI depends on scalable research-to-production pipelines. That belief fuels my active involvement in applied AI research, system design, and real-world experimentation. I’ve contributed to this field with four patents filed in areas such as generative AI orchestration, agentic system frameworks, intelligent feedback loops, and scalable ML infrastructure. Each invention reflects my focus on solving enterprise challenges in new and forward-looking ways.I’m also the author of “Continuous Machine Learning with Kubeflow: Performing Reliable MLOps”—a practical guide to building continuous learning systems in real-world production environments.I share my knowledge with professionals and enterprises seeking to make their ML pipelines scalable, compliant, and automation-ready.Having worked closely with data science, engineering, product, and executive teams, I understand what it takes to move from prototypes to platforms embedding compliance, explainability, governance, and ethical AI practices into every stage of the ML lifecycle.Today, I partner with organizations looking to create future-ready, AI-first infrastructures helping them unlock innovation while maintaining control, compliance, and competitive edge.If you’re building something ambitious in the AI space or want to evolve your current systems into intelligent, resilient platforms I’d love to connect and explore how we can build that future together
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.