Orchestrated the design and implementation of a Document-Based Question Answering system leveraging LangChain, Open AI LLM, and Pinecone to alleviate the workload for support teams with the MEVN stack. Experience in creating, modifying, and retrieving calendar events with the Google Calendar API to create meetings, using the 100.ms library to get script text of the conversation. Led the design and development of Prompt Engineering, including Summarization, Classification, Translation, Text Generation, and Question & Answering, to control output: length, style, audience, context, scenario-based guiding, chain of thought, and avoiding hallucinations. Researched Llama 2 and fine-tuned the Last Layer, Adapter Layers, Prefix Tuning, and LoRA Low Rank matrix train (less parameters, training time, RAM usage, switch between different LoRA weights), QLoRa (4 bit optimal) for privacy purposes and performance improvement. Built a chat application with the Open-Source LLM MPT-7B locally to protect privacy information for better customer answers, lower response times, and reduced costs. Successfully refactored complex code, reducing its complexity and enhancing readability by refactoring code from 5 embedded dimensions of if-else conditions to 1 dimension, implementing local processing instead of posting to the API server itself. Created reusable functions in the utils and services folder following SOLID, DRY, and KISS principles. Other developers failed to achieve this. Processed data for faster retrieval by using local storage, caching with Redis, storing files on S3, storing data in databases, and creating indexes for quick queries. Wrote background tasks to synchronize email data with the Database and S3 for quick retrieval, and wrote producers and consumers to exchange messages between microservices. Utilized Helm chart to dynamically inject variables for Kubernetes deployment. Used ORM Alchemy to create clean and readable code that is easy to switch to other databases such as MySQL and Postgres. Organized source code into services, models, utils, and controllers, which can be used across applications. Implemented testing, monitoring, debugging, reading code, reviewing code, communicating with other developers on design, guiding interns on coding, and writing documents.

Led the design and development of enterprise-level microservices, optimizing performance, driving $1 billion revenue every year. Created UML diagrams to visualize the classes in the projects. Used Composition over Inheritance to reduce deep, complex inheritance in classes and objects when developing new features. Efficiently managed ETL transaction data using PayPal and other financial instruments. Reduced API latency by up to 90% by optimizing queries in the database from 20-30 seconds to 1-2 seconds. Reviewed and provided feedback on PRs from other team members, ensuring code quality and efficiency.

Led the development of ERP applications for a prominent garment company, driving yearly revenue of $800 million using Django, Oracle DB, React, JQuery, and more. Designed and implemented comprehensive query reports for streamlined decision-making. Spearheaded the engineering of a dashboard page for the Board of Directors.

Contributed to the development of internet and mobile banking applications using Python, React JS, React Native, and Oracle, which drove a yearly revenue of $4 billion. Integrated multiple third-party systems to enhance functionality, resulting in a 20% increase in revenue. Managed front-end theming and layout building to ensure a seamless user experience.