Sanjay Mirani Email & Phone Number

San Diego, California, United States

• Developing YouDescribeX, a YouTube application catering to visually impaired users, leveraging MongoDB and PostgreSQL to ensure a robust and scalable database infrastructure.• Engineered and integrated a high-performance GPU pipeline using YOLO-V8 for object detection and custom AI models, significantly improving the efficiency of generating dynamic audio descriptions for video content.• Integrated YouTube API for seamless video retrieval and implemented RESTful APIs to connect frontend, backend, and GPU pipeline, enhancing accessibility for visually impaired individuals.• Designed an intuitive user interface with Wishlist and history features, prioritizing user experience and ensuring the application's accessibility in compliance with WCAG guidelines.• Technologies: React.js, Node.js, Python, Flask, MongoDB, PostgreSQL, YOLO-V8, OpenCV, NLTK, Pandas, FastAPI, Docker, AWS, GitLab CI/CD, Streamlit

San Diego, California, United States

• Developed Sign‑In Analyzer 3.0 (SA3) and managed Sign‑In Analyzer 2.0 (SA2) services, leveraging AWS OpenSearch Service (Elasticsearch), and implementing access restriction and data ingestion architecture, with expertise in back-end and database management systems. • Restructured and migrated SA3 service to remove corp host as part of Unfabric Migration, ensuring scalability and accessibility outside of Amazon’s Corp network. • Integrated AWS Cognito ‑ User Pool and Identity Pool with Federate in the service to enhance authentication and authorization functionalities. • Collaborated with an ANVIL security reviewer during the re‑architecture process, resulting in a 40% reduction in potential vulnerabilities and threats. • Migrated BotDetectionService (BDS) ingestion data from SA2 to SA3, utilizing TypeScript and CDK for lambda and Kinesis setup, facilitating efficient indexing into SA3 service.

San Francisco, California, United States

Tutored students in the construction of microscopes, taught them Python principles, and acquainted them with machine learning libraries such as Numpy, Pandas, OpenCV, and Matplotlib to collect and analyze pictures, as well as to design algorithms that turn visual information into scientific predictions.

Ahmedabad, Gujarat, India

• Collaborated on Traffic Governance System using Artificial Intelligence to optimize traffic flow and automated enforcement in urban cities.• Implemented Pool Surveillance System to detect potential drowning incidents using YOLO-v3 and Flask.• Utilized OpenCV for fast prototyping of computer vision tasks, enabling detection of incidents up to 8 seconds earlier. Actively collaborated with the team to debug and troubleshoot issues.