AI/ML & GenAI Engineer

Vihitha R

AI/ML Engineer

+1-857-***-**** ***********@*****.*** LinkedIn

SUMMARY

AI/ML Engineer with 4+ years of experience developing machine learning, NLP, and Generative AI applications across consumer technology and enterprise financial environments. Skilled in building recommendation systems, Retrieval-Augmented Generation

(RAG) platforms, predictive analytics models, and automated ML workflows using PyTorch, LangChain and BERT-based transformers. Proven track record of deploying production-grade AI services and scalable ML systems on AWS, Azure, and GCP, optimizing inference pipelines and working with large-scale structured and unstructured datasets across cloud-native infrastructure. SKILLS

Programming & Data: Python, SQL, PySpark, JavaScript, TypeScript, Shell Scripting, Git Machine Learning & Deep Learning: PyTorch, TensorFlow, Scikit-learn, XGBoost, BERT, RoBERTa, LLaMA, CNN, LSTM, Feature Engineering, Hyperparameter Tuning, Recommendation Systems, Anomaly Detection, Fraud Detection Generative AI & LLM Systems: LangChain, LangGraph, LlamaIndex, RAG Pipelines, HuggingFace Transformers, Prompt Engineering, LLM Fine-tuning, LLM Orchestration, Agentic AI Systems, Named Entity Recognition, Sentiment Analysis, Embeddings, Vector Search, Pinecone, ChromaDB, FAISS Data Engineering & Streaming: AWS Kinesis, PySpark, Spark, ETL Pipelines, Real-Time Streaming, Data Quality Validation Cloud & MLOps: AWS (S3, Glue, SageMaker, Kinesis, Redshift), Azure ML, Azure Synapse Analytics, GCP Vertex AI, MLflow, Docker, Kubernetes, CI/CD Pipelines, SHAP, Model Monitoring, LLMOps Databases: PostgreSQL, MongoDB, Redis, Neo4j, Snowflake, BigQuery, DynamoDB, Elasticsearch, OpenSearch APIs & Visualization: FastAPI, REST APIs, Power BI, Tableau, Grafana, Matplotlib, Plotly, Streamlit WORK EXPERIENCE

AI/ML Engineer Sony, USA June 2025 - Present

Engineered personalized recommendation systems using PyTorch, BERT-based content embeddings, and Redis caching to improve content discovery accuracy, increasing user engagement by 18% across digital media and streaming services.

Fine-tuned transformer-based NLP models including RoBERTa and LLaMA using Hugging Face and TensorFlow for multilingual sentiment analysis and intent classification on large-scale streaming interaction datasets.

Deployed scalable RAG-powered content search services using LangChain, LlamaIndex, and AWS SageMaker to support low- latency semantic retrieval and personalized content surfacing across streaming platforms.

Integrated PostgreSQL, MongoDB, and Neo4j graph-based relationship modeling with recommendation pipelines to enhance contextual ranking and personalized user experiences.

Automated ETL and feature engineering workflows using AWS Glue, Pandas, and NumPy, reducing data preprocessing time by 35% and improving model training efficiency.

Leveraged Claude Code to accelerate development of internal ML tooling and automation scripts, cutting engineering turnaround time on recurring pipeline tasks.

Optimized production ML pipelines through MLflow experiment tracking, hyperparameter tuning, CI/CD automation, and AWS CloudWatch monitoring, improving inference performance by 22% and deployment reliability across cloud environments. AI/ML Engineer Prudential Financial, USA Aug 2024 - April 2025

Evaluated customer policy and claims data with Scikit-learn and XGBoost models to identify retention patterns and improve risk prediction accuracy by 16%.

Processed large volumes of structured and unstructured insurance records through SQL, PySpark, and Azure Synapse Analytics to support enterprise reporting and underwriting analytics.

Applied Azure AI's Named Entity Recognition (NER) capabilities to extract key financial entities from policy documents and automate portions of document review workflows.

Built a regulatory compliance RAG pipeline using LangChain and ChromaDB for vector storage and Azure AI Search for enterprise-grade document indexing, classifying and cross-referencing policy language against regulatory requirements to reduce manual documentation lookup time by 30%.

Monitored model performance and deployment workflows through Azure Monitor, SHAP-based explainability reporting, and Azure DevOps CI/CD pipelines, reducing validation and deployment turnaround time by 28%, while using GitHub Copilot to speed up pipeline scripting and unit test coverage. ML Engineer Tata Consultancy Services (TCS), India Aug 2021 – Jul 2023

Built predictive analytics solutions with Python, Scikit-learn, and XGBoost to forecast customer behavior trends and improve business decision-making accuracy by 22% across enterprise reporting initiatives.

Consolidated large-scale transactional and operational datasets from PostgreSQL, MongoDB, and AWS Redshift environments to support data-driven forecasting and KPI analysis workflows.

Modeled classification and regression pipelines using Random Forests, SVM, and feature engineering techniques to strengthen customer segmentation and operational performance analysis.

Migrated legacy reporting workflows into automated ETL pipelines using AWS Glue, Pandas, and SQL, reducing manual reporting efforts by 40% across analytics teams.

Configured MLflow, Docker, and CI/CD pipelines to maintain model versioning, streamline deployment activities, and improve release efficiency for enterprise ML applications.

Generated interactive business intelligence dashboards through Power BI, Tableau, and Looker to monitor operational metrics, accelerating reporting turnaround time by 30% for cross-functional stakeholders. PROJECTS

Enterprise RAG Knowledge Assistant (LangChain, LlamaIndex, TypeScript, Azure OpenAI, Hugging Face, Pinecone, Redis, PostgreSQL, Docker, Azure Container Apps, MLflow)

Developed a RAG platform using LangChain for LLM orchestration and LlamaIndex for document indexing to enable contextual retrieval across enterprise documentation with a TypeScript-based frontend.

Structured semantic search pipelines using Pinecone for vector storage, Redis for caching, and PostgreSQL for metadata management to improve response relevance and retrieval efficiency for internal knowledge queries.

Containerized and tracked LLM workflows through Docker, Azure Container Apps, and MLflow to support scalable deployment and monitoring of AI-powered search applications.

Real-Time Recommendation Engine (PyTorch, Neo4j, Redis, AWS Kinesis, PySpark, PostgreSQL, Scikit-learn, Optuna, Collaborative Filtering)

Designed a real-time recommendation engine with PyTorch, Redis, and Neo4j graph modeling to personalize content ranking based on user interaction behavior.

Orchestrated streaming data pipelines using AWS Kinesis and PySpark to process high-volume engagement data for dynamic recommendation updates, storing processed outputs in PostgreSQL.

Tuned collaborative filtering and ranking models using PyTorch, Scikit-learn, and Optuna-based hyperparameter optimization to strengthen recommendation precision, improving click-through rate by 18%. MLOps Pipeline Automation (MLflow, Docker, Kubernetes, Jenkins, GCP Vertex AI, GCP Cloud Monitoring, CI/CD Pipelines)

Engineered automated ML deployment workflows using MLflow, Docker, Kubernetes, and Jenkins to streamline model versioning, container orchestration, and release management across multi-cloud environments.

Provisioned cloud-based training and inference environments on GCP Vertex AI to support scalable end-to-end ML pipeline execution and model serving across distributed infrastructure.

Monitored production model performance through GCP Cloud Monitoring, MLflow experiment tracking, and automated validation pipelines, improving deployment stability and reducing model degradation incidents by 20%. EDUCATION

Master of Science in Computer Science University of Massachusetts, Boston May 2025 CERTIFICATIONS

AWS Certified Machine Learning Engineer Associate

Microsoft Certified: Azure AI Engineer Associate

Contact this candidate