Krishna Singh Rajput

About

Data Scientist & Machine Learning Engineer

M.S. in Data Science @ ASU

Email: krishnas.r@outlook.com

City: Tempe, Arizona

Driven Data Scientist and Machine Learning Lead with a strong foundation in data science, machine learning, and cloud technologies. Led workshops for 100+ participants and developed AI-driven solutions, including an event recommender system that increased campus engagement by 30%. Known for quickly adapting to new tools and delivering results in complex problem-solving environments. Eager to apply my expertise to deliver impactful, data-driven insights.

Interests

Machine Learning

Statistical Analysis

Natural Language Processing - LLM

Big Data

Visualization

Data Pipelines

Deep Learning

Cloud Technologies

Education

Arizona State University, Tempe, AZ

Master of Science, Data Science and Analytics

2023 - 2025

Statisticas
Machine learning
Big Data Analytics
Natural Language Processing
Advanced Database Management

Pandit Deendayal Energy University, Gandhinagar, Gujarat

Bachelor of Technology, Mechanical engineering

2019 - 2023

Developed and deployed machine learning models for real-time bearing fault detection as part of a predictive maintenance solution utilizing ANN, XGBoost, and SVM

Experience

Research Project, ASU

AI Engineer

Sept 2024 - Present

Developed and implemented a Multimodal Large Language Model (LLM) using Vertex AI, Crew AI, and LangChain to enhance the accuracy of question-answering systems across various data modalities.
Integrated Google Gemini model via API to effectively process over 1,050 multimodal examples, demonstrating significant improvements over traditional models in accuracy and efficiency.
Led the creation of seven specialized AI agents, optimizing each for specific modalities such as text, images, and tables, to address complex challenges in Multimodal Question Answering (MMQA).

AI Society, ASU, Tempe

Vice President, Machine Learning Department

June 2024 - Present

Led a team of data scientists and machine learning enthusiasts, managing ML activities at the ML Lab, including project development and mentorship
Designed and executed a capstone project, providing participants hands-on experience in building end to end AI solution
Delivered engaging weekly workshops for over 100 participants on various ML topics to enhance skills knowledge
Hosted hackathons to promote innovation and teamwork, serving on the judging panel to evaluate and recognize outstanding AI projects

YourBeat Inc, Remote

Machine Learning Intern

May 2024 - July 2024

Collaborated with a team to design a chat-bot utilizing free-of-cost Large Language Models (LLMs) to assist music artists in generating personalized daily task lists aimed at professional growth
Conducted in-depth research on various available LLMs, Implemented text extraction, document chunking, and vectorization using LangChain, OpenAI embeddings, to assess their feasibility and integration within the project
Engaged in brainstorming sessions to refine the chatbot’s features and improve user engagement strategies
Contributed to the integration of LLM capabilities utilizing Retrieval augmented generation (RAG), enhancing its response generation based on user input

Projects

ASU Event Recommender System

Developed an event recommender system for ASU, utilizing Python and Beautiful Soup to extract data on over 3,000 upcoming events from ASU's website
Deployed web app using Flask, allowing users to input preferences and receive personalized event recommendations
Designed the entire data pipeline, employing machine learning techniques to analyze user preferences and suggest relevant events, enhancing campus engagement and user experience by 30%

Exploring Educational Outcomes

Analyzed the U.S. Department of Education’s College Scorecard dataset (3,000+ rows, 50+ features) to identify key factors influencing student success and inform data-driven strategies for educational institutions
Employed statistical techniques, including t-tests and ANOVA, using the SciPy library to perform hypothesis testing, uncovering five actionable insights

Wind Power Forecasting using ML

Conducted a comprehensive time series analysis to predict wind power on an hourly basis for a 15-day period
Utilized Long Short-Term Memory (LSTM) networks for accurate predictions, enhancing the model’s performance in capturing temporal dependencies in the data with R2 score of 82
Optimized maintenance scheduling, leading to an estimated 15% cost savings and improved resource utilization through strategic planning based on accurate forecasts

Bearing Fault Detection

Developed and deployed machine learning models for real-time bearing fault detection as part of a predictive maintenance solution to address bearing failure by classifying three types of bearing faults: inner-race, outer-race, and ball fault
Utilized ANN, XGBoost, and SVM, tuned model parameters to enhance performance, achieving effective anomaly detection

Sentiment Analysis

Performed sentiment analysis on over 1,40,000 Amazon reviews, accurately classifying sentiments as positive or negative
Implemented machine learning models including Naive Bayes and Perceptron, achieving an accuracy rate of 86% in sentiment classification
Utilized natural language processing (NLP) techniques for data preprocessing, enhancing model performance and delivering actionable insights for product improvements.

Central Banking Database System

The aim of this project is to develop and implement a comprehensive database system tailored to centralize the banking industry
This database, implemented using the Microsoft SQL Server database server, will include structured tables to manage critical information required for banking operations, covering areas such as bank details, ATM data, customer profiles, account records, transaction details, and card information