Mohan Hundre

About Mohan Hundre

I am a Senior GenAI Engineer with 8+ years of experience delivering production-grade Generative AI solutions on AWS and Kubernetes, with strong foundations in DevOps, CI/CD, and observability.Over the past 3 years, I have specialized in RAG pipelines and platforms:Built end-to-end workflows with Kubeflow & Katib for training, retraining, and hyperparameter tuning.Managed experiment tracking & model registry with MLflow to ensure reproducibility and auditability.Automated CI/CD & GitOps with Jenkins and ArgoCD, deploying models as microservices on EKS with KServe.Integrated Prometheus/Grafana for monitoring and Evidently AI for drift detection, ensuring reliability in BFSI environments.Ingested financial news & equity reports, chunked and embedded with OpenAI embeddings, stored in Pinecone, and deployed via FastAPI + GPT-4 microservice on EKS.Enabled analysts to query unstructured financial data and receive real-time contextual insights, reducing research turnaround by 77%.With proven expertise across banking, finance, and travel domains, I bring a strong combination of hands-on LLMOps engineering depth and business impact. I am passionate about building scalable, reliable, and explainable ML/GenAI systems that solve high-value problems.

Experience

1. Senior Genai Engineer

   Nagarro

   Jan 2023 — Present

   Architected and delivered a 3-agent Agentic AI system (Market Data Agent, Risk Analysis Agent, Report Agent) using LangGraph, GPT-4, and Python execution tools to automate Brevan Howard’s daily market-risk reporting workflow end-to-end.Developed enterprise-grade RAG pipelines using LangChain, OpenAI embeddings, and Vector database to retrieve internal risk rulebooks, exposure guidelines, and compliance policies for accurate limit validation and breach detection.Developed API-based market data pipelines with time-series validation for accurate VaR, volatility, drawdown, and correlation metrics.Designed Risk Analysis Agents capable of writing and executing Python code autonomously to compute financial risk metrics and generate actionable insights, reducing analyst workload by 90%+.Built report-generation and summarization agents using GPT-4 with grounded RAG context to produce analyst-ready PDF risk reports, charts, and teams alerts delivered to portfolio managers within minutes.Deployed the full Agentic AI pipeline as FastAPI-based microservices on AWS (EKS) with Docker, CI/CD automation (Jenkins), and environment-specific configurations.Established LLMOps/MLOps practices using MLflow, DVC, ArgoCD, and version-controlled risk rules to ensure reproducibility, auditability, and continuous evaluation of model and agent behavior.Implemented strict security guardrails including IAM role isolation, KMS encryption, VPC networking, and policy-aligned data flows to safely handle financial datasets, rulebooks, and internal exposure limits.Collaborated with portfolio managers, risk analysts, quants, and cloud architects to translate domain problems into scalable Agentic AI systems and mentored internal teams on RAG, Agentic AI, and LLM best practices.

Education

• Annamalai University, Annamalainagar

  Master of Business Administration - MBA, Information Technology

• University of Mumbai

  Bachelor's degree, Information Technology