• Took Product from prototype stage to production
• Curated Custom Dataset combining Open Source Datasets, Sources, and proprietary data
• Finetuned Pre-trained Object Detection and Action recognition Models on Custom Dataset using Distributed Data Parallel Training Approach with high Top-1 (87%), Top-5 Accuracy (95%), and F-1 Score (91%)
• Built End-to-End low latency (<1s) inference pipeline for real-time system with high throughput (25 CCTVs per node) for Weapon detection and Action Recognition (Violence) from live CCTV streams
• Deployed system using Nvidia Deepstream and Apache Kafka, integrating with Front-End Application providing Real-time Monitoring and Alerts

Big Data Analytics

• Supported students in understanding modern scale-out system architectures for processing large, structured and unstructured datasets, and how to scale resources for data processing based on workload requirements
• Provided hands-on support for students working with Airflow, Kafka, AWS, and Databricks to perform data analytics queries and implement data mining algorithms at scale
• Assisted students in Exploratory Data Analysis and Hypothesis Testing for Data Analysis

Database Design

• Helped students understand fundamentals of database management, relational database design, and data warehousing
• Guided students in designing and implementing relational databases, addressing challenges with large and heterogeneous datasets, such as transactional and geo-spatial data

Data Modelling and Storage

• Prepared presentations on topics such as Database Normalisation
• Provided technical assistance to students during sessions on SQLite and Python integration, addressing any questions

• Spearheaded the development of a Kiosk Information Management System, increasing operational efficiency and reducing manual inquiries by 40%
• Developed and deployed responsive, user-centric dynamic web applications using HTML, CSS, JavaScript, and React, enhancing client satisfaction and retention
• Conducted comprehensive code reviews and implemented rigorous testing protocols, reducing post-deployment defects by 25%