VARUN WOODI RAGHAVENDRA

About me

Hello!! I'm Varun W R, an ML & AI specialist with a passion for leveraging data to drive impactful solutions. I hold a Master's degree in Data Science from the University of Massachusetts, Dartmouth, where I graduated with a GPA of 3.97/4.0. My academic journey, combined with 3+ years of industry experience, has equipped me with a robust skill set to address complex business challenges using AI and machine learning technologies.

I specialize in frameworks like TensorFlow, PyTorch, and LangChain, along with cloud platforms such as AWS and Azure. My experience spans deep learning techniques, MLOps, LLM fine-tuning, and more. Throughout my career, I’ve consistently bridged the gap between research and product development, as demonstrated in my work on Multimodal LLM-powered medical image analysis.

With a strong foundation in machine learning and a passion for innovation, I’m here to help your business harness the power of AI to build the future. Let's connect and explore how we can turn data into your next competitive advantage.

Education

Master in Data Science

2022-2024

Bachelor of Technology in Electronics and Communication

2014-2018

Experience

Data Scientist

Healthi AI

Dec 2024 - Present

Collaborating with the team to enhance the Caddy Chatbot by integrating amazon.titan-embed-text-v2:0 for improved performance in the us-east-2 AWS region and refining chatbot logic using FastAPI.
Contributing to the design and integration of caddy-scraper with DynamoDB and PostgreSQL to streamline data storage and retrieval processes.
Assisting in developing scalable deployment pipelines using Docker and AWS, and documenting APIs with Swagger to improve usability and debugging.

Systems Engineer

Infosys

May 2021 - Aug 2022

Designed and ran experiments for time series data, improving forecast accuracy by 15%.
Contributed to projects aimed at enhancing pacemaker device safety by developing and testing Python scripts for real-time performance analysis. Improved testing efficiency by 30%.
Evangelized new machine learning technologies within the team and trained colleagues on adopting ML techniques.

Data Scientist

Varun Enterprises

Sep 2020 - Apr 2021

Spearheaded development and deployment of CNN models for thread nut anomaly detection, boosting revenue by 12%.
Transitioned custom CNN models from research into product using Flask, resulting in a 50% reduction in false positives.

Software Engineer

HCL Technologies

Dec 2018 - Dec 2019

Maintained and troubleshot trading applications, achieving a 95% resolution rate for production issues within 24 hours.
Implemented SQL queries for real-time data modifications, reducing system downtime by 20%.

Skills

Python
95%
TensorFlow, PyTorch
90%
Machine Learning
90%
SQL
95%
Statistics
90%
Machine Learning
95%
MLOps
85%
Deep Learning
90%
Power BI
80%
AWS
85%
Large Language Models
75%
Spark
80%

Projects

Multimodal LLM-powered Medical Image Analysis

Developed an advanced system using Multimodal-LLM to enhance diagnostic accuracy of MRI scans by providing textual and audio descriptions.

GitHub

Fine Tuning LLMs

Fine-tuned models like BERT, LLAMA, Mistral, and Zephyr using advanced techniques like PEFT, and optimized GPT-3.5 Turbo and GPT-4 with instruction fine-tuning and RLHF. Insights on these optimizations will be published soon.

GitHub

Streaming Data Pipeline with Kafka, Spark, and Advanced Visualizations

Built a big data pipeline for sourcing, processing, and visualizing data. Integrated Python for data sourcing, Kafka for streaming, Apache Spark for ETL processing, and Tableau for data visualization.

GitHub

End-to-End MLOps Project with ETL Pipeline - Network Security System

Developed ETL pipelines with MongoDB Atlas, optimized detection models with hyperparameter tuning and MLFlow tracking, and automated deployment using Docker, GitHub Actions, and AWS for real-time threat detection.

GitHub

View Other Projects

Blogs

Statistics

Intro to statistics

"Master the Foundations: Your Gateway to Data-Driven Insights."

Probability

"Unlock the Power of Uncertainty: The Heart of Statistical Thinking."

XGBoost Missing Data

"How XGBoost Handles Missing Data Efficiently: A Deep Dive"

Naive Bayes Algorithm

"Simplicity Meets Power: Fast and Accurate Predictions with Naive Bayes"

Dropout working

"How People Think Dropout Works vs. How It Actually Works"

Regularization in Deep Learning

"Preventing Overfitting: A Comprehensive Guide to Regularization in Deep Learning."

Fine-Tuning And Inferencing LLMs

"Fine-Tuning & Inference Excellence: Transforming LLM Capabilities"

LLM Application Dev: Prompt Templates

"Mastering LLM Application Development: Harnessing the Power of Prompt Templates"

Contact

## /content/drive/MyDrive/Colab Notebooks/Medical_Project/My_new_LLaVA/llava-ftmodel" BACK TOP

VARUN WOODI RAGHAVENDRA

About me

Education

Master in Data Science

Bachelor of Technology in Electronics and Communication

Experience

Data Scientist

Systems Engineer

Data Scientist

Software Engineer

Skills

Python

TensorFlow, PyTorch

Machine Learning

SQL

Statistics

Machine Learning

MLOps

Deep Learning

Power BI

AWS

Large Language Models

Spark

Projects

Multimodal LLM-powered Medical Image Analysis

Fine Tuning LLMs

Streaming Data Pipeline with Kafka, Spark, and Advanced Visualizations

End-to-End MLOps Project with ETL Pipeline - Network Security System

Blogs

Statistics

Intro to statistics

Probability

Machine Learning

XGBoost Missing Data

Naive Bayes Algorithm

Deep Learning

Dropout working

Regularization in Deep Learning

Large Language Models

Fine-Tuning And Inferencing LLMs

LLM Application Dev: Prompt Templates

Contact