Jameer Babu

Experience

• 

  Amrita Centre for Wireless Network and Applications

  Aug 2018 - Dec 2018

  Student

  Working on Delay Tolerant Networks using ONE simulator for simulations. Studied protocols like Epidemic, PRoPhet, Spray and Wait. Developed existing Binary spray and wait protocol by adding a new parameter called priority.

• 

  Computational Intelligence - Course Work

  Oct 2018 - Nov 2018

  Classification of Boys and Girls using Linear Regression

  Data consisting the features of girls and boys was saved in .csv file. Gradient Descent Algorithm is applied with Mean square error as cost function. Weights are obtained with a success rate of 99.97℅

• 

  Ziroh Labs

  Jun 2019 - Jul 2020

  Data Science Intern

  Worked on building backend system for a Federated Machine Learning project basedon Client - Server model which can be analogous to Keypad suggestion system where mobilebeing client and Recommendation System being Server, with main emphasis on Privacyand Security using ML libraries such as Tensorflow, Pytorch and Keras.Models were trained, validated, tested with data being distributed between server with50 percent and rest with clients. Formalized relationship between the weights obtained from server and client. Fully Homomorphic Encryption is used for securing and Approximation functions werederived for activation functions using Mathematical empirical assumptions with similaritiesbetween plots of activation and approximation functions .Client model weights and Server model weights were averaged according to the distribution.Results with accuracy, network transmission speed were calculated on MNIST, CIFARdatasets Show less

• 

  OvalEdge

  Sept 2020 - Jul 2022

  Governance Engineer

  • Saved over 1000-person hours by implementing a computer vision-powered Data Governance tool to ensure privacy, utilizing Image recognition and detection algorithms to identify and classify sensitive elements, such as faces, for privacy protection.• Leveraged NLTK and word embeddings (CBOW, Skip-Gram) within the governance tool to automate metadata extraction, categorization, and tagging, improving data discoverability and reducing manual overhead by 80%• Developed machine learning models, including KNN, K-Means, SVM, Naïve Bayes, and Random Forest, within the governance platform to monitor and predict data quality issues, resulting in a 30% reduction in data errors. Show less

• 

  Edvenswa Enterprises

  Aug 2022 - Jul 2023

  Software Engineer

  • Reduced compute costs by 40% in deep learning models for inventory prediction and demand forecasting within the Warehouse Management System (WMS) by leveraging historical data, applying LSTM and RNN architectures for time series forecasting, and optimizing load-based scaling and batch processing.• Led the development of AI-powered features for WMS, driving 11+% growth in operational efficiency by integrating CNN for visual data processing and leveraging Autoencoders, GANs and VAE for data augmentation and generation to improve inventory optimization and real-time tracking. Show less

• 

  Northwest Missouri State University

  Jan 2024 - Dec 2024

  Graduate Research Assistant

  Designed RAG-enabled Chatbot for Northwest admission department using Large Language Models (LLMs), OpenAI models, Vector DBs, FAISS (Facebook AI Similarity Search), Knowledge Graphs, Hugging Face and LangChain, improving text comprehension accuracy by 95%Built RAG-enabled Chatbots and prompt-based AI agents, processing over 10,000 documents daily and reducing query response time by 70% using AWS SageMaker, VisionAI, Streamlit, Google Collaboratory.