Milind Choudhary Email & Phone Number

Austin, Tx, Us

Dallas, Texas, United States

Developing NLP/ML-based models to assist companies in targeting customers in the financial sector

Dallas, Texas, United States

Bias Detection in Social & Moral Values in COVID-19 and Immigration Tweets: • Orchestrated a data transformation pipeline and implemented feature engineering to generate a new dataset of 1.5 Million tweets to represent the tweets as Frames of Communication and cross-validated using RLHF• Finetuned GPT3.5 & GPT-4o using LORA to identify moral and cultural values and analyze the rationalesAI Chatbot for Spanish Corpus• Created a Spanish language corpus focused on political conflict using PySpark and developed a framework to finetune different LLMs for developing a Spanish chatbot after statistical analysis and visualization using MatplotlibApproximate Counting Via hypergraphs based on Problog and Graph Neural Network• Developed and applied a new approximate counting in Statistical Learning & compared it with benchmarks

• Designed a MultiAgent-based AI framework for GE Vernova’s Distribution system by using LLAMA-3 and Mistral 7B• Developed an AI/ML-powered fault simulation system that triggers applications based on scenarios (e.g., storms) and generates NLP reports to reduce operator workload after pre-processing data using Pandas and Scikit-learn• Implemented ML models to detect anomalies in Hadoop File System Logs with the highest accuracy of 99%

Dallas, Texas, United States

• Developed a super-resolution framework for single source images and video for better detection of license plates and house numbers using CNN models like Inception Resnet & GoogleNet and surpassed the benchmark of 92% by 4%

• Deployed the first Container as a Service (CaaS) version of InfoScale (Efficient Storage Technology with low-latency backup and robust security) with the help of Kubernetes and enhanced the UI/UX on Openshift Console• Augmented the functionality of InfoScale by adding disk-group information feature to reduce accessibility time of information by 1 minute (100% improvement) and streamlined deployment using CI/CD pipelines• Developed a debugging tool for Erasure Coded Volumes and integrated it with InfoScale for data security

Dublin, Ireland

• Developed a library for rapid development of optimized Deep Learning models to be used through CLI, and improved the accuracy of the models on Diabetes, and Genetic Database by 5%

• Performed ETL on Satellite data and applied ML & ANN models to improve the resolution of precipitation from (5X5 sqkm ) to (1X1 sqkm) and augmented accuracy from 65% to 85%

Master Of Science - Ms, Computer Science, 3.86

Song Recommender based on Prolog and LLMs • Built a song recommender system utilizing the causal behavior of interrelated songs.

Bachelor Of Technology, Computer Science, 3.4

Activities and Societies: President of Coding Club (SPAWN), Part of organizing team of Technical and Cultural festivals, Class.

High School, Science

Activities and Societies: Academic Excellence, Football team and Debating Team