GenAI & Machine Learning Full-Stack Engineer

Kasturi Duggaraju

+1-571-***-**** *******.*****@*****.*** LinkedIn

SUMMARY

Full-Stack GenAI / Machine Learning Engineer with 4 years of experience designing, developing, and deploying production-grade Generative AI and Machine Learning applications across Azure and AWS. Proven expertise in building end-to-end Retrieval-Augmented Generation (RAG) systems, LLM-powered AI agents, semantic search platforms, conversational AI applications, and intelligent automation solutions using GPT-4 (Azure OpenAI & AWS Bedrock), LangChain, FastAPI, Python, React, and Next.js. Experienced across the complete AI application lifecycle, including document ingestion, document chunking, embedding generation, vector search, semantic retrieval, prompt engineering, function calling, API development, cloud deployment, Kubernetes orchestration, CI/CD, monitoring, and MLOps. Strong background in developing scalable, secure, cloud-native AI solutions with Azure AI Search, Pinecone, Qdrant, and FAISS, delivering measurable improvements in application performance, retrieval quality, and operational efficiency. TECHNICAL SKILLS

• Programming Languages: Python, TypeScript/JavaScript, Java, SQL, C++

• Generative AI & LLMs: GPT-4 (Azure OpenAI, AWS Bedrock), LangChain, RAG, AI Agents, Prompt Engineering, Function Calling, Hugging Face Transformers, Semantic Search, Conversational AI, Knowledge Retrieval, Vector Embeddings, Document Intelligence

• Machine Learning: Machine Learning, NLP, Deep Learning, Scikit-Learn, TensorFlow, PyTorch, XGBoost

• Backend & APIs: FastAPI, Flask, REST APIs, WebSockets, gRPC

• Frontend: React, Next.js

• Vector Databases & Data Storage: Azure AI Search, Pinecone, Qdrant, FAISS, Amazon OpenSearch, PostgreSQL, MySQL, MongoDB, DynamoDB

• Cloud Platforms: Azure OpenAI, Azure AI Services, Azure Machine Learning, Azure Functions, Azure Blob Storage, Azure SQL, Azure Key Vault, AKS, Azure DevOps, AWS Bedrock, SageMaker, Lambda, API Gateway, Amazon EKS, Amazon S3, Amazon RDS, AWS Glue, Athena, Amazon Redshift, AWS Cognito

• MLOps & DevOps: Docker, Kubernetes, MLflow, GitHub Actions, Azure DevOps, AWS Code Pipeline, CI/CD, Git, GitHub, GitLab

• Monitoring & Observability: Azure Monitor, Application Insights, Amazon CloudWatch, Prometheus, Grafana

• Security & Responsible AI: Azure AD B2C, Azure Key Vault, AWS IAM, AWS Cognito, Responsible AI, Data Privacy, Audit Logging

EXPERIENCE

Gen AI Engineer at Walgreens Feb 2025 – Present

Designed, developed, and deployed production-grade GenAI applications using React, TypeScript, FastAPI, Python, and Azure OpenAI (GPT-4) to support enterprise document intelligence, conversational AI, and knowledge management use cases.

Architected and implemented Retrieval-Augmented Generation (RAG) solutions using LangChain, Azure AI Search, Azure Blob Storage, vector embeddings, and semantic retrieval, enabling employees to query enterprise knowledge repositories using natural language.

Built scalable document ingestion pipelines supporting PDF and enterprise document processing, including parsing, chunking, embedding generation, metadata enrichment, indexing, and retrieval optimization to improve search relevance.

Developed LLM-powered AI agents using GPT-4 Function Calling and LangChain tools to automate multi-step workflows such as report generation, enterprise information retrieval, and business process assistance.

Engineered real-time conversational AI experiences using WebSockets with token streaming, improving application responsiveness and user interaction for enterprise chatbot solutions.

Built AI-powered analytics applications integrating Azure SQL with GPT-4, enabling conversational insights and natural language querying over structured enterprise datasets.

Improved answer quality and reduced hallucinations by optimizing retrieval strategies, prompt engineering, grounding techniques, and context management across RAG pipelines.

Reduced application latency by 35%, lowering p95 response time from 1.8 seconds to 1.2 seconds through API optimization, intelligent caching, and efficient retrieval architecture.

Deployed containerized AI services on Azure Kubernetes Service (AKS) using Docker and automated CI/CD pipelines through GitHub Actions, improving deployment consistency and release efficiency.

Implemented monitoring, logging, and observability using Azure Monitor and Application Insights to ensure production reliability and rapid issue detection.

Strengthened enterprise security through Azure AD B2C, Azure Key Vault, audit logging, and Responsible AI governance to support privacy, compliance, and secure LLM usage.

Collaborated with product managers, security teams, and business stakeholders to deliver scalable, production-ready AI applications aligned with enterprise governance standards. Python ML Engineer at Accenture Aug 2022 – Dec 2024

Designed and developed LLM-powered enterprise applications using GPT-4 (AWS Bedrock), LangChain, FastAPI, and Python for document summarization, semantic search, enterprise virtual assistants, and intelligent knowledge retrieval.

Built production-ready Retrieval-Augmented Generation (RAG) pipelines using FAISS, Amazon OpenSearch, Amazon S3, embeddings, and semantic retrieval, enabling context-aware responses across enterprise document repositories.

Developed semantic search capabilities using Pinecone and Qdrant vector databases to improve similarity search, contextual retrieval, and knowledge discovery.

Designed scalable AI inference architectures using AWS Lambda, Step Functions, and API Gateway, reducing processing latency by 40% while supporting event-driven LLM workflows.

Built full-stack AI applications using React, TypeScript, FastAPI, PostgreSQL, and AWS cloud services, delivering responsive user experiences and scalable backend APIs.

Managed machine learning workflows on Amazon SageMaker, including model training, experiment tracking with MLflow, deployment, and monitoring to support production ML lifecycle management.

Developed scalable ETL pipelines using AWS Glue, Athena, PySpark, and Amazon Redshift to prepare enterprise data for machine learning and AI applications.

Containerized and deployed AI applications on Amazon EKS, implementing Kubernetes autoscaling, load balancing, monitoring, and high-availability deployment strategies.

Supported platform scalability during 5 traffic growth (120 RPS) through Kubernetes-based autoscaling and infrastructure optimization, improving system reliability and throughput.

Collaborated with DevOps teams to automate deployments using Docker, CodePipeline, CloudFormation, and CI/CD, reducing deployment effort and improving operational consistency.

Implemented Responsible AI practices focusing on model governance, audit logging, enterprise data privacy, and secure deployment of AI systems.

PROJECTS

AI Confessor – Conversational AI Platform

Developed a real-time conversational AI platform using GPT-4, FastAPI, React, and WebSockets, delivering low-latency multi-turn conversations through streaming responses.

Integrated Vosk for speech-to-text capabilities and deployed scalable Docker-based microservices to support concurrent user interactions.

On-Demand Professor Q&A Bot

Built an LLM-powered question-answering platform using GPT4All and Qdrant, enabling semantic search and context- aware responses over domain-specific knowledge.

Deployed Qdrant using Docker and implemented vector indexing to support efficient similarity search and scalable retrieval. EDUCATION

Bachelors in Electronics and Communication Engineering, Jawaharlal Nehru Institute of Technology

Contact this candidate