SAI SUDHA MOOLA

I blend engineering, creativity, product management, and design thinking. My technical expertise and strategic outlook drive innovative, user-centric solutions. I seek to harmonize data, insights, vision, and strategy to steer products and projects.

• Runner in the 2022 IBC Hackfest in the 'Blockchain and WEB 3.0' hackathon conducted by the International Blockchain Congress.
• L1 Team Lead at Street Cause VNRVJIET

Airplane Object Detection in Ultra-low-Resolution Satellite Images

[Presented at the International Conference on Advances in Emerging Trends in Computer Applications (ICAECT-2023)] 

Tech Stack: Python (Keras, Tensorflow), Colab/Jupyter Notebook, PyTorch IDE, CNN, YOLOv8, Object Detection

Developed a YOLOv8-based aircraft detection model for satellite images, using a custom dataset of 1000+ augmented images due to a lack of structured datasets. The model's lower computational cost than YOLOv5 makes it suitable for real-time applications. 

Real-Time Data Analytics Platform on AWS

Tech Stack: Amazon Kinesis, AWS Lambda, DynamoDB, AWS QuickSight

Created a real-time data analytics platform utilizing Amazon Kinesis to handle streaming data, such as IoT sensor readings or stock prices. Implemented parallel processing by employing Kinesis Shards and AWS Lambda for distributed processing and scalability. Conducted real-time analytics and visualized the results using AWS QuickSight. 

Information Accessibility on Websites by Visually Impaired People

Tech Stack: HTML, CSS, Bootstrap, JavaScript ES6, Speech Recognition API, DJANGO

An online food delivery website for visually impaired users, using speech recognition for easy navigation and voice output to read website content aloud. 

Data Processing and Transformation in Hive using Azure VM 

Tech Stack: Microsoft Azure, Hadoop, Apache Hive, Bash Scripting, HDFS

Developed a scalable data processing and transformation system using Apache Hive on an Azure VM. Provisioned an Azure VM, setting up a Hadoop ecosystem, and using Hive for large dataset processing. Key tasks included data ingestion into HDFS, creating external tables in Hive, and executing complex queries for data analysis and transformation. By the usage of parallel processing techniques, the project optimized data workflows to handle big data efficiently, resulting in a reliable and scalable system adaptable to varying workloads.