Vedansh Gupta
Data scientist by day, algorithm whisperer by night. Turning caffeine into code and decoding the secrets of data. Occasionally baffled by data's whimsy 🤖💻.
Python
C++
C
MySQL
Cassandra
StreamLit
Pandas
Numpy
SCIPY
KERAS
Scikit-learn
TensorFlow
Flask
Docker
GIT
Amazon AWS
Heroku
DJango
Jupyter
NoSQL
Analyzed the data thoroughly and predicted whether the employee gets promoted or not with a accuracy of 92% from Decision trees
===============================================
Skills: Matplotlib · Scikit-Learn · XGBoost · Gradient Boosting · NumPy · Machine Learning · Pandas · Seaborn · Data Analysis · Random Forest
I developed an end-to-end ML project to predict wine quality with a user-friendly UI for input. I fetched, validated, preprocessed, and cleaned data from an AWS S3 bucket, then used various algorithms, selecting RandomForestRegressor for its 96% accuracy. I integrated MLFlow and Dagshub for experiment tracking, performed hyperparameter tuning with GridSearchCV, and deployed it as a Docker container on an AWS EC2 instance using GitHub actions for CI/CD.
A deep learning project which predicts whether the person has kidney tumor or not. Has an accuracy of above 94%.
===============================================
Skills: OOPS · JSON · GitHub · Dagshub · Flask · Deep Learning · TensorFlow · DVC · Amazon Web Services (AWS) · Modular Programming · MLflow · Git · Convolutional Neural Networks (CNN) · Python (Programming Language) · Dockers · Image Processing
Engineered a relational database for 10+ tables using MySQL to optimize book store operations. Crafted SQL queries and stored procedures for efficient sales.Engineered a relational database for 10+ tables using MySQL to optimize book store operations. Crafted SQL queries and stored procedures for efficient sales.
===============================================
Skills: MySQL · XAMPP · SQL · Data Analysis · Database Management System (DBMS)
An end to end Conversational QnA Chatbot, designed to provide intuitive and accurate responses to user queries across a wide range of topics, leveraging the advanced NLP techniques and the state-of-the-art capabilities of OpenAI's language model. Used Streamlit for smooth User Interface and huggingface for deployment to production.
===============================================
Skills: Generative AI · LangChain · OpenAI · Streamlit · Huggingface
ATS built using Google Gemini, streamlit and huggingface. The link to it is attached. So, just go upload a job description of the company to which you are applying for and get your resume score, your blunders & more. Used Streamlit for smooth User Interface and huggingface for deployment to production.
===============================================
Skills: Generative AI · Gemini · Streamlit · Huggingface
Ever wished you could get the gist of a YouTube video without watching the whole thing? Our cutting-edge app, powered by the Gemini-pro LLM, does just that! Simply input the link to any YouTube video, and watch as our AI delivers a concise, engaging summary in seconds. Perfect for quick learning, research, or staying up-to-date with the latest trends—get all the key points without the time commitment. Try it now and revolutionize the way you consume video content! 📺✨
===============================================
Skills: Generative AI · Gemini · Streamlit · Huggingface
It helps you save your time by reading your invoices and answering your questions such as discounts, due amount, deadlines and much more... I've attached the link to my project. Visit and chat with your invoices.
===============================================
Skills: Generative AI · Gemini · Streamlit · Huggingface
An AI nutritionist Using Google Gemini Pro Vision, a large image model with the help of streamlit and google gemini pro api. The LLM app will read the uploaded image given to it & after converting it into bytes will tell all the nutrients, pros and cons of having the food that is displayed in the image.
===============================================
Skills: Generative AI · Gemini · Streamlit · Huggingface
DocConverseGemini is an LLM application built using Google gemini pro model and langchain where we can query multiple pdf documents with the help of FAISS vector emebeddings and can fetch any information present inside the PDFs.
===============================================
Skills: Generative AI · Gemini · Large Language Models (LLM) · LangChain · Streamlit · Huggingface · FAISS
End to end RAG app using streamlit... Took a webpage : https://lilianweng.github.io/posts/2023-06-23-agent/... Scraped it using beautifulsoup python package... Converted to vectors using huggingface embeddings... Stored it in Cassandra vector database, and queried the results from the db
===============================================
Skills: Retrieval-Augmented Generation(RAG) · AstraDB · Vector Databases · bs4 · Cassandra · LangChain · Streamlit · Groq · Hugginface-embeddings · Mixtral
Here, I used Objectbox VectorDB, a new and efficient way of vector embeddings, with OpenAI's newly launched GPT-4o(omni) model... Used GPT-4o(omni) model from OpenAI... Took US_cencus PDFs for querying... For word embeddings, used OpenAIEmbeddings...Used ObjectBox VectorDB stores for storing vectors... Finally, created retrieval chain and invoked it for answering the query.
===============================================
Skills: Vector Databases · ObjectBox · Retrieval-Augmented Generation(RAG) · LangChain · Streamlit · openai-embeddings · gpt-4o
Drop me a mail at vedanshgupta606@gmail.com
Or call me at +91 9305959634
Built with ❤️ and dev.voyage