Nirnay Reddy Gaddam
Boston,USA
Summary
AI / Agentic AI Engineer with 5 years of experience building enterprise-scale AI platforms, LLM applications, and distributed ML systems across insurance, life sciences, banking, and retail domains. Specialized in agent-driven AI workflows, Retrieval Augmented Generation (RAG), real-time inference architectures, and enterprise knowledge systems. Experienced deploying production AI services on AWS/Azure using vector search, microservices, and event-driven pipelines while maintaining strong standards for observability, reliability, and data governance in regulated environments.
Overview
7
7
years of professional experience
Work History
AI / Machine Learning Engineer
Liberty Mutual Insurance
Boston, MA
07.2025 - Current
- Architected an enterprise LLM-driven claims intelligence platform used by operations teams to automate document summarization, policy interpretation, and fraud signal extraction across insurance workflows, serving hundreds of internal users and thousands of claims documents daily.
- Designed agent-based AI orchestration workflows coordinating LLM reasoning, knowledge base retrieval, and rule validation for underwriting and claims decision pipelines.
- Implemented enterprise RAG architecture integrating policy documentation and claims knowledge with vector search (FAISS / Pinecone), improving retrieval accuracy by ~18%.
- Built multi-step reasoning AI agents capable of invoking enterprise tools and querying structured data sources, reducing manual claims document review time by ~30%.
- Engineered real-time AI inference architecture integrating LLM services, APIs, and enterprise data systems, lowering AI response latency by ~25%.
- Developed event-driven pipelines using Azure Event Hub enabling real-time claims event ingestion and automated inference triggering.
- Built ground-truth evaluation pipelines using labeled datasets to benchmark LLM responses and monitor hallucination rates.
- Implemented observability pipelines using MLflow and cloud monitoring, enabling detection of model drift and performance regressions across production AI services.
- Developed internal R2D2-style retrieval services allowing AI agents to securely access enterprise knowledge bases and claims datasets.
- Exposed AI capabilities via FastAPI microservices, integrating with underwriting platforms, Puma request workflows, and internal claims systems.
- Implemented LLM guardrails and output monitoring while integrating outputs with the Enterprise Analytics Platform (EAP) for downstream analytics and operational insights.
- Designed platform security controls including RBAC, secret rotation, API authentication, and secure token management for Vertex services.
Machine Learning Engineer
Charles River Laboratories
Wilmington, MA
06.2024 - 06.2025
- Built an AI-assisted biomedical research platform enabling automated literature summarization and semantic retrieval using LLM pipelines.
- Implemented vector search infrastructure (FAISS) supporting semantic retrieval across biomedical research repositories.
- Designed LLM-powered research assistants capable of retrieving insights from large-scale scientific knowledge bases.
- Engineered distributed PySpark pipelines processing genomic datasets across cloud platforms.
- Delivered ML inference APIs via FastAPI, enabling researchers to query predictive models and AI assistants.
- Deployed containerized ML workloads using Docker, Kubernetes, and CI/CD pipelines for scalable model deployment.
- Managed ML lifecycle operations including experiment tracking and monitoring using MLflow.
Data Scientist
BBVA (Accenture)
Hyderabad, India
06.2021 - 05.2023
- Developed credit risk and fraud detection models trained on large financial transaction datasets.
- Built real-time fraud analytics pipelines using Spark and Kafka processing millions of events.
- Implemented NLP pipelines analyzing customer complaints and call center transcripts to identify fraud signals.
- Delivered fraud scoring services via REST APIs using Flask and FastAPI for real-time risk evaluation.
Data Scientist
Electronics Mart India Limited
Hyderabad, India
04.2019 - 05.2021
- Built demand forecasting models improving inventory planning across large retail networks.
- Engineered ETL pipelines using Python, SQL, and Spark processing sales and inventory datasets.
- Developed customer segmentation models supporting targeted marketing and recommendation systems.
Education
Master of Science - Computer Science
Northeastern University
Massachusetts12-2025
Skills
- ML / AI: PyTorch, TensorFlow, Scikit-learn,RAG, Prompt Engineering, Agent Workflows
- Data Engineering: Spark, PySpark, Kafka, Azure Event Hub, Pandas, NumPy
- Vector & Databases: FAISS, Pinecone, PostgreSQL, MySQL, MongoDB, Snowflake
- Cloud & Infrastructure: AWS (EC2, S3, SageMaker), Azure ML, Docker, Kubernetes
- MLOps: MLflow, Experiment Tracking, Monitoring, CI/CD
- Security: RBAC, Secret Rotation, API Authentication, Secure Key Management
- Languages: Python, SQL
- APIs & Services: FastAPI, Flask, REST APIs
Timeline
AI / Machine Learning Engineer
Liberty Mutual Insurance
07.2025 - Current
Machine Learning Engineer
Charles River Laboratories
06.2024 - 06.2025
Data Scientist
BBVA (Accenture)
06.2021 - 05.2023
Data Scientist
Electronics Mart India Limited
04.2019 - 05.2021
Master of Science - Computer Science
Northeastern University
Core Expertise
Agentic AI Systems - Enterprise LLM Applications - Retrieval Augmented Generation (RAG) - AI Workflow OrchestrationTool - Using AI Agents - Knowledge Base Engineering - Vector Search Infrastructure - Event-Driven AI Systems - Real-time AI Inference - Distributed Data Pipelines - AI Guardrails - Enterprise AI Deployment