About

I'm pursuing Master of Science in Information Systems at Northeastern University, Boston. Are you seeking a dynamic and challenging role where you can leverage your diverse skill set and passion for innovation? Look no further! With a solid foundation in data analysis, machine learning, software development, and a proven track record of delivering impactful solutions, I am ready to tackle new challenges and make a meaningful contribution to your team. With experience spanning across diverse industries and technologies, I bring adaptability, creativity, and a strong drive for excellence to every project. Let's connect and explore how my expertise can drive success in your organization.

After pursuing Co-Op at Novelis, I am ready to pursue my NEXT EXCITING JOURNEY!

I'm open for full-time opportunities starting January 2024!!

Versatile Data Professional | ML & AI Aficionado | Software Engineer.

  • University: Northeastern University
  • Degree: Master of Science in Information Systems
  • City: Boston, USA
  • Email: varsha.hindupur@gmail.com
  • GitHub
  • LinkedIn
  • Medium Blog Profile & Articles
  • Courses:
  • 1. Advances in Data Science and Architecture (INFO7390)
  • 2. Application Engineering and Development (INFO5100)
  • 3. Big Data Systems and Intelligent Analytics (INFO7245)
  • 4. Data Science Engineering Methods and Tools (INFO6105)
  • 5. Generative Artificial Intelligence (AI) (INFO 7375)
  • 6. High Parallel Computing in Machine Learning & AI (CSYE7105)
  • 7. Program Structure and Algorithms (INFO6205)
  • 8. AI/ML Prompt Engineering (INFO7375)

Skills

MLOps
Backend Python
Docker, Kubernetes, Airflow
Machine Learning & Data Science
Cloud Architecture & Computing
Database
Frontend
Data Engineering

Professional Experience

Sumary

Varsha Hindupur

Driven and creative ML Infused Software Engineer with a strong background in Data Science and MLOps, leveraging Python and SQL for insightful data modeling, visualization, and statistical analysis. Pursuing Master of Science in Information Systems at Northeastern University, Boston, with a passion for developing data-driven strategies and solutions.

  • Boston, MA
  • varsha.hindupur@gmail.com

Education

Master of Science in Information Systems &

2022 - 2024

Northeastern University, Boston, MA

I am a Data Science aficionado currently pursuing Master of Science in Information Systems at Northeastern University. My keen interest lies in developing data-driven strategies, blueprints, and solutions, with a specialization in Data Science & Machine Learning Operations. Proficient in Python and SQL, I possess a deep understanding of Data Modeling, Visualization, and various statistical paradigms.

Bachelor of Engineering in Information Technology &

2012 - 2016

K.J.Somaiya (KJSIEIT), Mumbai, Maharashtra, India

Throughout my Bachelor's journey, I have honed my expertise in areas such as programming, database management, web development, networking, and cybersecurity. With a passion for innovation and problem-solving, I am dedicated to leveraging my knowledge to create efficient and cutting-edge solutions in the field of Information Technology.

Professional Experience

Graduate Research Assistant (Software Engineering)

Feb 2024 - May 2024

Accounting Department

Northeastern University at Boston, MA



  • Skills: Route53, S3, MongoDB, FastAPI, ReactJS, Mongoose Framework, AWS QuickSight Data Analytics

R&D Ecosystem Co-ordinator Co-Op

May 2023 - Dec 2023

R&D Ecosystem & Data Science

Novelis - An Aditya Birla Company at Atlanta, GA



  • Skills: Power BI, Salesforce, LLM, Arize, Azure MLOps, KPI monitoring

Graduate Research Assistant (Machine Learning)

Feb 2023 - May 2023

Department: Marketing Research Academia

Northeastern University, Boston, MA



  • Skills: · Computer Vision, CNN, DeepFace, Image Analysis, AWS Sagemaker, HPC (Discovery)

HandCraftedStyles

Jan 2020 - May 2022

Co-Founder

Mumbai, India

  • Led marketing, driving 50K+ INR in sales in a year through content creation and customer engagement.

Senior Quality Analyst - Automation

Dec 2020 - Aug 2022

Group: Credit Reporting Agency

UST Global - Xpanxion (Blueconch Technologies), Pune, India

  • Client: US Top Credit Reporting Company


  • Skills: GCP, AWS, Databricks, Hadoop, Terraform, Snowflake, Collibra, Salesforce

Senior Software Test Engineer Analyst

Mar 2019 - Dec 2020

Accenture, Mumbai, India


  • Skills: ETL, Restful APIs, AWS (EMR, S3, Redshift), PySpark, SQL, Stored Procedures, UI/UX, Jira

Test Engineer

Mar 2017 - Mar 2019

Client: Banking

Infosys, Pune, India



  • Skills: Core Java, React, JavaScript, Automation Testing, NoSQL, GraphQL

Game Development Intern

Nov 2016 - Jan 2017

Client: In-House Game Development

Rendered Ideas, Mumbai, India



  • Skills: Sentimental Analysis, Microsoft Excel, SAS, ETL

Projects

Presenting a collection of my diverse undertakings – a carefully curated assortment of projects that showcase my dedication and creative abilities.


FoodKing Interactive: LLM Gemma-Powered Dining Experience

June 2024

Tech Stack: Streamlit, Ollama Gemma:2B, Docker, Python

  • Developed a chatbot to assist customers in placing orders and answering queries about the FoodKing restaurant menu.
  • Utilized Streamlit for building the web interface and Ollama Gemma:2B for natural language processing, all running locally using Docker.
  • Implemented a dynamic menu system with selectable sizes and quantities, integrating the chatbot for seamless interaction.
  • Achieved a user-friendly interface where customers can view the menu, interact with the bot, and manage their cart simultaneously.
Github

Predict Lung and Colon Cancer

Apr 2024

Tech Stack: Discovery (NEU HPC), ResNet50, Neural Network Modeling

  • Trained, executed ResNet50 on 5000 image dataset having 5 cancer types, producing results in 36 seconds with 90% accuracy.
  • Methods used to increase the performance of the models: Serial Execution, DDP (Distributed Data Parallel) on CPU, DDP on GPU, Model Parallelism, AMP (Automatic Mixed Precision)
  • And, as per the result part we could train our model in 30 seconds using the parallel processing techniques and still maintain the accuracy to 93% overall.
  • The procedures used were 24 times efficient when compared to Serial Executions on CPU. Speedup received was more than 50%. Overall, if all the model executes in 30 seconds with the use of modern compute resources, then we have achieved our goal of making the product available to consumers and businesses.
Github

Netflix Movie Recommendation Engine

Jan 2024 - Present

Tech Stack: Data Mining, Data Science, Data Visualization, Machine Learning

  • Added innovative 'Mood Based Recommendations' reflecting users' emotional states.
  • Cleansed, refined, and integrated Netflix Prize Data from TXT files into a single CSV containing user ratings, customer IDs, movie IDs, titles, dates, genres, and release years.
  • Utilized web scraping to extract genre information from IMDb for analysis and recommendation.
  • Implemented Collaborative Filtering to suggest movies based on users' watch history.
  • Executed code on Northeastern's High Performance Computing Virtual Machine Discovery for 50% surge in performance.
Github

SpaceX Data Science & Analysis Project

Tech Stack - Python, Data Science, Machine Learning, Exploratory Data Analysis (EDA)

Github

  • This Data Science project explores SpaceX's operational successes, focusing on First Stage landings, cost management, and time efficiency.
  • Analysis of First Stage landings involves studying mission data, Falcon 9 and Falcon Heavy rocket technologies, and precise landing algorithms.
  • Cost management scrutiny includes SpaceX's use of reusable components, innovative manufacturing, and strategic procurement, along with partnership assessments.
  • Time efficiency analysis examines project management methodologies, scheduling strategies, and automation use.
  • The project aims to provide actionable insights for SpaceX's mission goals and strategies.


    • Real-Time Data Streaming - A Data Engineering Project

    Tech Stack - Apache Airflow, Python, Apache Kafka, Apache Zookeeper, Apache Spark, Cassandra, PostgreSQL, Docker

    Github

  • Implemented end-to-end real-time population count data streaming pipeline at Northeastern University.
  • Utilized a powerful stack of tools and technologies, including Apache Airflow, Python, Apache Kafka, Apache Zookeeper, Apache Spark, and Snowflake.
  • Containerized the entire pipeline using Docker for efficient deployment and management.


    • Ride Price Forecasting (Uber/Lyft)

    Tech Stack - Linear Regression, Decision Tree, Gradient Boosting Regressor, Folium

    Github

  • Analyzed Kaggle data on Uber availability, car preferences, and price-weather correlation, employing binning for surge multipliers and conducting RFE for attribute correlations with cab fare prices.
  • Utilized Linear Regression, Decision Tree, and Gradient Boosting Regressor, achieving high accuracy (MAE, MSE, RMAE values closer to 1) in price predictions.


    • Text Analysis - Natural Processing Language

    Tech Stack - Python: NLTK, SpaCy, Sci-kit Learn, Gensim, TF-IDF, Seaborn, SQLite

    Github

  • Conducted a thorough analysis of 50,000 IMDB movie reviews using Natural Language Processing, Neural Networks, and Computational and Semantic data modeling to uncover dominant themes and sentiments, providing valuable insights into customer preferences and guiding strategic business choices.
  • Tech Stack: Python, NLTK Libraries (spaCy, PyTorch, Tensorflow)
  • Models: Text Ranking, Linear Discriminant Analysis (LDA), KMedoids, Neural NetworksMachine Learning 📈 Conducted a thorough analysis of 50,000 IMDB movie reviews using Natural Language Processing, Neural Networks, and Computational and Semantic data modeling to uncover dominant themes and sentiments, providing valuable insights into customer preferences and guiding strategic business choices. Tech Stack: Python, NLTK Libraries (spaCy, PyTorch, Tensorflow) Models: Text Ranking, Linear Discriminant Analysis (LDA), KMedoids, Neural Networks
  • Skills: Data Visualization · Python (Programming Language) · PyTorch · Natural Language Processing (NLP) · Statistical Modeling


    • Aircast - Air Quality Prediction

    Tech Stack - Python, Streamlit, FastAPI, Airflow, AWS RDS, AWS EC2, Hugging Face LSTM ML Model

    Github

  • 🖥️ Designed and implemented the front-end of this web application using the Streamlit library in Python.
  • 🔬 Established a Data-as-a-Service solution hosted in the cloud, which involves data scraping from AWS cloud, followed by extensive Data Engineering. This data was then stored in Amazon RDS for easy accessibility.
  • 📊 Developed a Long Short Term Memory (LSTM) Neural Network for Air Quality forecasting and hosted it as a standalone Machine Learning service in the cloud, achieving an impressive RMSE value of 0.8.
  • ⚙️ Crafted daily Airflow DAGs for data scraping and LSTM model execution while utilizing NGINX for load balancing.
  • 🔍 Constructed APIs via FastAPI in Python to facilitate interactions between the aforementioned services.
  • 🕰️ Integrated all the services using Docker, enabling hosting on various platforms, including AWS.


    • Titanic Survival Prediction using Neural Network

    Tech Stack - Python, NumPy, Pandas, Matplotlib, Logistic Regression

    Github




    MeetIn - a Meeting Intelligence Application

    Tech Stack: Python, Streamlit, FastAPI, Airflow, AWS RDS, AWS EC2

    Github

    2023

  • 💡Developed a sophisticated voice-to-text-image pipeline, seamlessly transcribing audios and videos into text and images. Implemented intelligent query responses through the integration of GPT-3.5 Turbo, enhancing interactive capabilities.





    • Regency, Frequency, Monetory - KMeans Clustering

    Tech Stack: Python, Kmeans Clustering Machine Learning Algorithm

    Github

  • 💡 Performed Data Wrangling to clean the datasets and remove null, incorrect values
  • 📈 Calculated Regency, Frequency, Monetary; Recency tells us how many days since the last transaction for each customer, frequency tells you how frequently a customer shops and, monetary tells you the total shopping spending for each customer
  • 🏬 These are the business aspects for an online store (completely generic)
  • 🖥️ Implemented KMeans clustering model from Scikit-learn package and calculated the value of K by Elbow Method and Silhouette Method
  • 📈 Identified Elbow method to give the best K value for this particular model
  • ⚙️ Performed KMeans clustering algorithm with Scipy package and executed this manual implementation of the ready-made model available in Scikit-Learn
  • 🔍 Verified the K value for the cluster and 3 was found the best fit

    • Medium Blog





    Exploring Power BI's Analytical Prowess: Unleash Your Data's Secrets
    Medium Link




    Bringing Your Streamlit Web App to Life on Azure App Services
    Medium Link




    LeetCode & HackerRank Adventures: Conquering Advanced SQL and Windows Functions for Database Mastery
    Medium Link




    Summarizing lecture from Data+AI World Tour by Databricks: Delta Live Tables A to Z: Best Practices for Modern Data Pipelines
    Medium Link

    Certifications

    LinkedIn link to all my Certificates!


    Extra-Curricular

    Proud to be a "Tech Expert" at @NEUBlockChain student organization

    Hobbies

    Fitness, Yoga, Reading, Sports, Hiking, Art, Music, Theatre

    Contact

    Let's Connect and Collaborate! I'm thrilled to hear from you and explore potential opportunities. Whether you have a project in mind, a question to ask, or just want to say hello, feel free to get in touch. I'm here to engage in meaningful conversations and exciting ventures.


    varsha.hindupur@gmail.com


    hindupur.v@northeastern.edu