GPU SCheduling Optimization

• This project develops a multi-objective optimization model for scheduling jobs on heterogeneous GPU clusters using real workload data from the Alibaba cluster trace.
• (https://github.com/deepa-cv/GPU_Scheduling_Optimization)

ADHD Task Manager

• This project infers the ADHD subtype of the person using Gemini-2.5Pro and displays an adaptive UI accordingly. Built using React.js and Gemini API. Deployed on Vercel
• (https://github.com/apurvab15/adhd-task-manager)

Reward Hacking detection in Language Models

• This project aims to separate reward hacked responses from normal responses during inference. Trained on a custom dataset generated by altering the reward model using GRPO and the Big-Math dataset.
• (https://github.com/josephzhong/cs762_project)

Catalog Indexing Engine

• This project employs microservices based architecture and supports both structured and unstructured queries to retrive information from fashion catalogs, displaying the most relevant products to end users, with the aid of machine learning techniques
• (https://github.com/BFB-PES/Catalog-Indexing-Engine)

Predicting Code Runtime Complexity using Static Analysis and Machine Learning

• The aim of this project is to predict the Big O Notation of a given algorithm in three different languages, C++, Java, and Python using Abstract Syntax Trees(AST), Graph Embeddings, Machine learning based classification algorithms and BiLSTM.
• (https://github.com/Capstone-B10-2022)

Kinship Recognition using Siamese Networks

• This project aimed to analyze whether two people are related by blood or not through their facial pictures using Siamese Networks. Tech included Python, Tensorflow 2.0, and Matplotlib
• (https://github.com/Pravena725/Kinship-Recognition)

COVID Data Analysis and Information Retrieval

• This project aimed to recommend new insights to bolster the fight against COVID by generating new insights to keep up with the acceleration of literature related to the disease
• (https://github.com/deepa-cv/AIWIR_project)

Predicting Student Assessment Marks to Prevent Student Attrition in MOOCs

• The aim is to predict the student's score, rather than just predicting if he/she will pass or fail the course using ensemble learning techniques.
• (https://github.com/deepa-cv/Prevent-Student-Attriton)

Deploying CRUD operations based Flask Application using Microservices based Architecture

• Docker and Kubernetes are almost synonymous to 'microservices' as they help package and manage the different components of a project/ application, thereby easing up the implementation of a microservices architecture.
• (https://github.com/076-096-105-106-CloudHack-1/CloudHack_1)

San Francisco Crime Classification using Pyspark

• This project aimed to classify crime using incremental Learning and streaming techniques to train models based on new data encountered using Python, and Pyspark.
• (https://github.com/deepa-cv/Crime-Classification-on-streaming-data )