Omkar Kadam

Background

About Omkar Kadam

I am a Senior Software Engineer with 4+ years of experience sitting at the intersection of Generative AI and Cloud-Native Architecture. My passion lies in building intelligent, scalable systems that don't just "run" but actively solve complex business problems through automation and machine learning.Currently, I focus on integrating Large Language Models (LLMs) into enterprise workflows to automate diagnostics and enhance user decision-making, while maintaining a strong foundation in AWS Serverless architecture and backend optimization.Here is a breakdown of my technical journey and core competencies:GENERATIVE AI & LLM ENGINEERING:I specialize in deploying production-grade AI solutions that drive operational efficiency. RAG Pipelines: Architected a diagnostic assistant using AWS Lex and OpenAI. Engineered a Retrieval-Augmented Generation (RAG) pipeline to fetch real-time data from MongoDB and AWS S3, reducing manual support inquiry time by 35%. Context Optimization: Implemented context management strategies using Mongo Memory Store to handle conversation history, reducing API token usage and redundant calls by 20%. Tooling: Experienced with LangChain, AWS Bedrock, and Hugging Face for building agents that can reason and execute tasks.CLOUD ARCHITECTURE & SERVERLESS:My core background is in modernizing legacy systems into highly available, event-driven architectures. Serverless Migration: Led the migration of an asset delivery pipeline to a serverless stack (AWS Lambda, S3), eliminating manual steps and reducing delivery cycle time by 70%. Event-Driven Orchestration: Designed the "FastTrack" fulfillment system using AWS Step Functions, reducing processing time by 80%. Performance: Built media inspection services capable of analyzing attributes in under 20 seconds (previously 5+ minutes).BACKEND OPTIMIZATION:I am committed to writing efficient code that saves time and money. Cost Reduction: leveraged AWS DynamoDB (Single Table Design) to slash storage costs by 60% and query latency by 90%. Traffic Optimization: Developed logic to detect redundant asset deliveries, cutting data transfer costs by 40%.TECHNICAL TOOLBOX AI/ML: OpenAI API, AWS Bedrock, AWS Lex, LangChain, LangGraph, RAG, Prompt Engineering, Vector Databases. Cloud (AWS): Lambda, Step Functions, SQS, SNS, API Gateway, DynamoDB, S3, EC2, IAM. Languages: Python, Java, SQL, Shell Scripting. DevOps & Tools: Docker, Kubernetes, Terraform, Git, Jenkins, Jira.