Skip to content
View sidde95's full-sized avatar

Block or report sidde95

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sidde95/README.md

Siddhartha Roy

Designing Data Intelligence Systems — from Machine Learning to Generative AI

📄 View My Resume


About Me

I build intelligent data and AI systems that merge structured learning with generative reasoning.
My work focuses on Machine Learning, LLM Architectures, and AI workflow automation using frameworks like LangGraph and CrewAI.

This space documents my experiments — transforming models, agents, and pipelines into interactive, real-world applications.


Core Expertise

  • Applied Machine Learning — Regression, Classification, Feature Engineering
  • Generative AI Systems — LangChain, LangGraph, CrewAI, RAG Pipelines
  • Deep Learning & NLP — TensorFlow, Keras, NLTK, Transformers
  • Deployment — Streamlit-based AI and ML applications

Featured Repositories

End-to-End Generative AI projects LLM-driven systems that automate reasoning, summarization, and analytics.

  • AI Ticket Triage System (LangGraph) – Automates ticket classification and routing
  • Business Consultant Analyst (CrewAI) – AI agent generating business insights from datasets
  • QueryMyPDF (RAG) – Conversational interface for document understanding

Stack: LangChain · LangGraph · CrewAI · Pinecone · ChromaDB · FAISS · OpenAI API · HuggingFace API · Groq API · Streamlit


Exploring Recurrent Neural Networks (RNNs) for sequential data analysis and sentiment classification.
This project focuses on understanding text emotions using RNN architectures trained on Twitter and Reddit sentiment datasets.

  • Twitter and Reddit Sentiment Analysis – Performs sentiment classification (Positive, Neutral, Negative) using Recurrent Neural Networks.

Stack: TensorFlow · Keras · NLTK · pandas · numpy · matplotlib · seaborn · Streamlit


End-to-end Deep Learning projects built with TensorFlow and Keras showcasing how ANNs solve real-world problems in both regression and classification tasks.

  • Spotify Churn Prediction – Classification using Artificial Neural Networks (Streamlit Deployed)
  • Beats Per Minute (BPM) Prediction – Regression model predicting song tempo (Kaggle Playground Series - S5E9)

Stack: TensorFlow · Keras · scikit-learn · pandas · numpy · matplotlib · seaborn · Streamlit


End-to-end ML workflows for predictive modeling and data-driven decision systems.

  • Medical Premium Price Prediction – Regression model for healthcare pricing
  • Bank Churn Prediction – Predicts customer churn using ensemble ML
  • Rainfall Forecasting – Gradient boosting for weather prediction
  • Used Car Price Prediction - Predict the selling price of used cars using features such as brand, manufacturing year, kilometers driven, etc.

Stack: scikit-learn · Linear Regression · Logistic Regression · Decision Tree · AdaBoost · Random Forest · XG Boost · Gradient Boost · KNN · pandas · numpy · matplotlib · seaborn · Streamlit


Tech Stack


Research & Learning Path

  • Advancing LangGraph-based LLM architectures for autonomous reasoning
  • Building CrewAI agents for data analytics and business intelligence
  • Deepening study in RNNs, Transformers, and Large Language Models
  • Experimenting with Generative AI pipelines for real-world data systems

Pinned Loading

  1. GenerativeAI-Projects GenerativeAI-Projects Public

    Python

  2. ANN-Projects ANN-Projects Public

    Jupyter Notebook

  3. Machine-Learning-Projects Machine-Learning-Projects Public

    Jupyter Notebook

  4. RNN-Project RNN-Project Public

    Jupyter Notebook