Hi, my name is

I'm a Data Scientist.

I'm passionate about transforming complex data into actionable insights. I specialize in machine learning, statistical analysis, and data visualization to solve real-world problems and drive data-driven decisions.

Check out my projects!

About Me

Hello! I'm Nishant Bharti, a passionate Data Scientist with expertise in machine learning, statistical analysis, and data engineering. I specialize in transforming complex datasets into actionable insights that drive business growth and innovation.

With experience in both industry and research, I've had the opportunity to work on diverse projects ranging from predictive modeling and natural language processing to computer vision and big data analytics. My approach combines strong technical skills with business acumen to deliver data-driven solutions that solve real-world challenges.

Currently, I'm focused on leveraging advanced machine learning techniques to extract meaningful patterns from data and build intelligent systems that make a tangible impact. I'm particularly interested in the intersection of AI and business strategy, where data science can drive innovation and create competitive advantages.

Here are some of the technologies and tools I work with:

  • Python (Pandas, NumPy, Scikit-learn)
  • Machine Learning
  • Deep Learning (TensorFlow, PyTorch)
  • Data Visualization (Matplotlib, Seaborn, Plotly)
  • SQL & NoSQL Databases
  • Big Data (Spark, Hadoop)
  • Natural Language Processing
  • Computer Vision
  • Cloud Platforms (AWS)
  • Operating System Development
Nishant Bharti

Where I've Worked

Co-Founder @ IRLVibes

Dec 2024 - Present

  • Co-founded IRLVibes, an innovative online thrift store that revolutionizes second-hand shopping through gamification and AI-driven personalization.
  • Developed a unique points and rewards system that encourages sustainable shopping behaviors, increasing user engagement by 65% and average session duration by 3x.
  • Implemented AI-powered recommendation engines that suggest personalized thrift finds based on user preferences, browsing history, and social interactions.
  • Designed interactive shopping challenges and treasure hunts that drive user retention and create a community around sustainable fashion.
  • Integrated social features that allow users to share finds, create wishlists, and participate in virtual clothing swaps, fostering a vibrant community of eco-conscious shoppers.

Data Scientist @ HiLabs

July 2022 - May 2024

  • Designed and implemented an end-to-end clinical summarization pipeline with Named Entity Recognition (85% precision, 77% recall) for medical entities including Diagnosis, Procedures, and Medications, enhancing clinical data processing.
  • Developed automated systems for PDF Medical Chart processing, achieving 79% accuracy in DRG code prediction, resulting in $1.2M cost savings through 90% reduction in fraud and over-billing.
  • Engineered a Text-to-Code solution predicting ICD10 and SNOMED CT codes with 83% accuracy, significantly improving clinical data standardization and decision-making processes.
  • Led the development of NLP models including fine-tuned BERT for clinical sentiment analysis (86% accuracy) and YOLO v5 for medical document structure recognition (75% precision, 80% recall).

Data Scientist Intern @ HiLabs

Feb 2022 - Mar 2022

  • Developed and implemented a robust pipeline to automate provider directory updates and maintenance, leveraging multiple data sources and algorithms to fetch the latest information, score directory accuracy, and suggest updates.
  • Augmented the provider directory with multidimensional provider information, including demographics and affiliations, to provide a comprehensive view of the providers, enabling data-driven decision-making and enhancing the overall quality of the directory.
  • Created Python scripts for cleaning and processing extracted data into structured formats, ensuring data quality and consistency for analysis.

Education

M.Tech in Computer Science and Engineering

Indian Institute of Technology Guwahati

2020 - 2022

  • Specialized in Data Science and Machine Learning and Hardware Compression and Encoding Architecture
  • Relevant coursework: Machine Learning, Deep Learning, Data Structures and Algorithms, Computer Networks

B.Tech in Computer Science and Engineering

MRSPTU, Bathinda

2015 - 2019

  • Relevant coursework: Data Structures, Algorithms, Database Systems, Computer Architecture, Computer Networks, Operating Systems

Research

ZOCHEN: Compression using Zero chain elimination and encoding to improve endurance of NVMs

Published in: ISQED'23 (International Symposium on Quality Electronic Design)

ZOCHEN is a novel compression and encoding technique that enhances the endurance of Phase-Change Memory (PCM) in Non-Volatile Memory (NVM) systems. By efficiently eliminating zero-value bit chains and implementing fine-grained encoding with minimal tag bits, ZOCHEN significantly reduces bit-flips in PCM cells. Our approach demonstrates substantial improvements in memory lifetime and performance compared to existing methods, offering a practical solution to the write endurance challenges in NVM-based main memory systems.

View Paper

Some Things I've Built

Featured Project

TabForest (Ongoing)

A browser extension that transforms your browsing into real-world environmental impact. For every 1000 tabs opened, TabForest plants a real tree, turning your digital habits into positive environmental action.

  • JavaScript
  • Chrome Extension API
  • HTML5
  • CSS3

Featured Project

Project OS (Ongoing)

An educational journey into operating system development, building a custom OS from the ground up. This project explores low-level system programming, memory management, process scheduling, and hardware interaction, following the principles of modern OS design while learning from wyoos.org and osdev.org.

  • C++
  • Assembly
  • x86 Architecture
  • QEMU

Featured Project

Crypto Dashboard

An interactive cryptocurrency dashboard that displays real-time market data, price charts, and trends for various cryptocurrencies. Built with Streamlit, it provides users with up-to-date market insights and analytics.

  • Python
  • Streamlit
  • Pandas
  • Plotly

Other Noteworthy Projects

view the archive

Speech Recognition using VQ

An implementation of speech recognition using Vector Quantization with K-Means and LBG algorithms. The system processes speech signals into 12-dimensional feature vectors and generates efficient codebooks for pattern matching.

  • C++
  • Signal Processing
  • Machine Learning

Election Fraud Detection using Benford's Law

An analytical tool that applies Benford's Law to detect potential anomalies in election results. The application analyzes the frequency distribution of leading digits in vote counts to identify statistical irregularities that may indicate electoral fraud.

  • Python
  • Pandas
  • Matplotlib

Image Cryptography using RSA

An implementation of image cryptography using RSA encryption. The system converts images into pixel values and applies RSA encryption to secure the data.

  • Python
  • RSA

Get In Touch

What's Next?

I'm always looking for new opportunities, my inbox is always open. Whether you have a question or just want to say hi, I'll try my best to get back to you!

Say Hello