About Me

Experienced Machine Learning Engineer with 4+ years in AI, software development, and collaboration. Adept at building complex neural networks, automating ETL/ML pipelines, and deploying innovative AI solutions. Proven success in optimizing project efficiency, reducing candidate evaluation time by 50%, and leading projects resulting in increased product sales and user satisfaction. Skilled in Python, Java, NodeJS, React Native, and various databases.

My Resume

Skills

Here are some of my skills

Machine Learning/Artificial Intelligence

  • Tensorflow
  • Pytorch
  • Keras
  • HuggingFace
  • Langchain
  • OpenAI

Programming

  • Python
  • Java
  • NodeJS
  • React-Native

Frameworks

  • FastAPI
  • Flask
  • Springboot
  • Java Play
  • React

Databases

  • MongoDB
  • Elastic Search
  • MySQL
  • Redis
  • PostgreSQL

Cloud

  • Git, Github, Bitbucket
  • AWS
  • Azure
  • Docker
  • Kubernetes

Others

  • Kafka
  • Airflow
  • Storm
  • Data Visualalization

Projects

Movie Recommendation System

In this project, I created a movie recommendation system that can suggest relevant and personalized movies to users based on their search. I used Python as the programming language, MongoDB as the vector database, and text embeddings from Sentence Transformer. I also implemented two different similarity metrics: cosine similarity and euclidean distance. The movie recommendation system can provide relevant and personalized suggestions to users.

Movie Recommendation System

Tic Tac Toe Game

This project is a classic implementation of the Tic-Tac-Toe game built using the Pygame library. The game provides an engaging experience with a user-friendly interface for playing against an AI opponent. The AI opponent utilizes a strategic search algorithm called MinMax to make intelligent moves, providing a challenging match for players.

Tic Tac Toe Game

Question Answering Application

This project utilizes BERT Language model from Hugging face and Pinecone vector Database to search from the given documents and provide relevant answers to the user queries.

Question Answering Application

Image Analysis and Generation

This project has Image Analysis and Image Generation functionalities developed using the Microsoft's Azure service.

Image Analysis and Generation

News Category Classification

This project aims to develop the best machine learning model to accurately categorize news articles based on the headline and short description text. The data came from a Kaggle dataset containing over 200,000 articles labeled into 42 categories.

News Category Classification

RAG based QnA

This project is a Document based Question Answering utilizing the langchain for document reading, parsing and hugging face for creating embeddings and vector stores for storing these embeddings and using the corresponding Large Language model for retrieving the relevant documents from the vector store and providing the answers to the user queries.

RAG based QnA

Personalized Conversational Assistant

In this project I have utilized OpenAI apis to to finetune gpt-3.5-turbo model on my personal whatsapp chat data and created a model which can reply to the messages as I would have replied to them.

Personalized Conversational Assistant

Trend Prediction

This project is based on a complex neural network using LSTM’s to identify the trends in stock price movement and predict the trend of the stock, then using statistical analysis on the predicted data and make a suggestion of “BUY”, “SELL” or “HOLD”. The model performed exceptionally well and accurately predicts the trend and makes a strong suggestion

Stock Trend Prediction