● Led successful migration of legacy Ruby on Rails application to AWS Cloud, implementing comprehensive cloud infrastructure using EKS, ECR, Lambda, and other AWS services. ● Designed and implemented database-centric metering system using PostgreSQL stored procedures, ensuring scalability and real-time usage tracking. ● Designed and implemented secure authentication system using AWS Cognito with multi-factor verification. ● Developed ETL pipeline using AWS Kinesis Firehose, processing and analyzing large-scale data streams. ● Created smart contracts for NFTs and custom tokens, implementing secure reward systems and airdrop functionality.

● Led Experiments for Few Shot Learning using LLM. Benchmarked models including GPT-4o, GPT-4o-mini, Mixtral-8x7B v0.1, Llama 3.1 8B, and Llama 3.1 70B, achieving 81% accuracy with GPT-4o and successfully architecting production deployment. ● Engineered data ETL for Language Model training, achieving 10x speed improvement and ensuring data integrity. ● Lead document processing optimization initiative, achieving 3x performance improvement and 9x cost reduction through architecture optimization. ● Architected and implemented Salesforce integration, developing custom APIs and leveraging Salesforce Objects and Metadata API for seamless data synchronization. ● Engineered internal reporting system reducing report generation time from hours to minutes, enabling real-time data analytics and decision-making. ● Delivered client-specific document extraction solutions achieving 90%+ accuracy through custom ensemble of Deep Learning Models, LLMs, and proprietary frameworks. ● Developed and implemented key-value extraction framework reducing API development cycle by 70%.

● Developed ML models for voice disease classification achieving 80%+ accuracy using CNN and traditional ML algorithms. ● Implemented automated data collection system using Selenium, reducing manual effort by 90%.