Developed and deployed enterprise-grade Generative AI solutions for clients across multiple industries, delivering real-world business value through applied AI.

Built end-to-end AI applications enabling document intelligence and knowledge discovery, supporting faster and more informed decision-making.

Collaborated with cross-functional and client-facing teams to deliver scalable, production-ready solutions aligned with organizational goals.

• Collaborated with a team of 5 developers to build and maintain Admin, Client, and Clinician portals, working on both frontend

(Angular.js) and backend (Node.js, LoopBack) while following Agile methodologies on Azure DevOps.

• Developed and enhanced RESTful APIs using Node.js and LoopBack, integrating with MongoDB to meet client requirements.

• Managed and resolved user stories on Azure Boards, ensuring timely completion of sprint deliverables while using Git for version

control.

• Developed comprehensive full-stack portals for USC - Alzheimer’s Therapeutic Research Institute (ATRI), utilizing ReactJS for the frontend and Django for the backend with PostgreSQL, ensuring responsive design and scalable architecture.

• Implemented Test Driven Development (TDD) and Agile methodologies, employing Cypress for frontend testing, Selenium for backend testing, and Jira for task management to achieve high reliability and performance.

• Created a CI/CD pipeline using GitHub Actions to automate frontend, backend, and functional tests before deployment, enhancing the overall development workflow and deployment efficiency.

• Developed a LLM Text2SQL model to generate and execute SQL queries on custom databases using the GPT-3.5-turbo model and Azure OpenAI embeddings, incorporating client-specific information and business rules for improved quality

• Led backend development with tools such as FastAPI, Pydantic, and SQLAlchemy, and managed the team codebase using Azure DevOps for efficient collaboration and version control

• Engineered a CI/CD pipeline for seamless deployment of the bot on Azure Cloud, integrating it with MS Teams, implementing Single Sign-On, and enhancing the chat UI with TypeScript.

• Created a federated learning model using PySyft and PyTorch to improve the privacy and security of the model. This model was able to detect anomalies in Bitcoin transactions by leveraging the power of federated learning.

• Employed Truffle to develop a custom dataset with transactions through Ethereum and track all attackers. This involved working with blockchain technology to collect and store transaction data for analysis.

• Utilized Tensorflow to train several Deep learning models such as SVM, Autoencoder, variational autoencoder, and deep belief network for anomaly detection in Bitcoin transactions. These models were tested and optimized to improve the accuracy of anomaly detection.