I'm a Data Scientist focused on Machine Learning and AI, with hands-on experience building complete, deployable applications rather than stopping at notebooks.

My work spans the full lifecycle of data and ML systems — from SQL-based analysis and data preprocessing to model development, evaluation, explainability, modular inference pipelines, deployment, and production-oriented application design.

I'm particularly interested in Deep Learning, Computer Vision, NLP, Generative AI, RAG systems, and MLOps.

• 🔭 Building end-to-end ML & AI applications
• 🧠 Working with Machine Learning, Deep Learning, Computer Vision & GenAI
• 🚀 Turning trained models into interactive and deployable applications
• 🧩 Focusing on modular, maintainable, and reproducible project architecture
• 📚 Continuously expanding into advanced AI engineering and MLOps

Python SQL Pandas NumPy SciPy

Scikit-learn XGBoost Feature Engineering Model Evaluation SHAP Imbalanced Learning

PyTorch TensorFlow CNN RNN LSTM GRU Deep Learning

OpenCV Pillow Ultralytics YOLO Object Detection Image Classification ONNX

NLP LangChain RAG Retrievers Chains Embeddings Vector Search LLM Applications Ollama OpenRouter

Matplotlib Seaborn Plotly Power BI Tableau Excel

Streamlit Flask FastAPI Docker Git GitHub MLflow DVC Pytest

MySQL MongoDB FAISS Pinecone

End-to-end computer vision application for detecting and localizing visible vehicle damage from images.

Highlights

• YOLO11n object detection
• 6 vehicle damage categories
• Bounding-box localization and confidence filtering
• Modular inference pipeline
• Structured prediction handling
• Precision, Recall, mAP@50 and mAP@50–95 evaluation
• Streamlit deployment
• Centralized configuration, logging, exception handling and testing

Tech: Python PyTorch Ultralytics YOLO11n OpenCV Pillow Streamlit

🚀 Live Demo  •  💻 Repository

A modular Retrieval-Augmented Generation system combining document ingestion, embeddings, FAISS vector search, retrievers, chains, an LLM backend, and a Streamlit frontend.

Highlights

• Retrieval-Augmented Generation pipeline
• Hugging Face inference embeddings
• FAISS vector store
• Retriever + document chain architecture
• LangChain integration
• FastAPI backend
• Streamlit frontend
• Structured logging, validation and error handling
• Docker and cloud deployment configuration
• Test suite and modular application architecture

Tech: Python LangChain FastAPI FAISS Hugging Face OpenRouter Streamlit Docker LangSmith

🚀 Live Demo  •  💻 Repository

Production-oriented image classification system built with PyTorch and a custom AlexNet architecture.

Highlights

• 10-class image classification
• Custom PyTorch model architecture
• Modular training and inference workflow
• ONNX inference support
• Interactive Streamlit application
• Reproducible deep learning workflow

Tech: Python PyTorch AlexNet ONNX Streamlit Computer Vision

🚀 Live Demo  •  💻 Repository

Deployment-ready time-series forecasting application using a Vanilla RNN trained with PyTorch to forecast household electricity consumption.

Highlights

• Time-series preprocessing and feature engineering
• 24-hour sliding-window forecasting
• Vanilla RNN implemented in PyTorch
• Inference-only deployment pipeline
• MLflow experiment metadata
• Modular src/ architecture
• Multi-page Streamlit application
• Reusable pretrained model and scaler artifacts

Tech: Python PyTorch RNN Pandas Scikit-learn MLflow Streamlit Plotly

🚀 Live Demo  •  💻 Repository

Complete Machine Learning system combining SQL analytics, EDA, feature engineering, model experimentation, prediction pipelines and Streamlit deployment.

Highlights

• SQL-based customer analysis
• End-to-end preprocessing pipeline
• Classification model comparison
• XGBoost and Scikit-learn
• SHAP explainability
• Reusable model artifacts
• Streamlit prediction application
• Production-style project structure
• MLOps-oriented tooling including MLflow and DVC

Tech: Python SQL Pandas Scikit-learn XGBoost SHAP Streamlit MLflow DVC Docker

🚀 Live Demo  •  💻 Repository

My GitHub also includes projects covering:

• Customer Segmentation using Unsupervised Learning
• Fraud Detection
• Deep Learning-based Student Academic Outcome Prediction
• Human Activity Recognition with a Deep Residual MLP
• Retail Demand Forecasting
• Retail Data Warehouse & Analytics
• Sales Business Intelligence Dashboard
• Insurance Cross-Sell Analytics with NoSQL
• Market Basket Analysis & Product Recommendation
• NLP and Generative AI applications
• Data Analysis and Machine Learning portfolio projects

👉 Explore all repositories:
https://github.com/GouravGC?tab=repositories

I'm continuing to deepen my expertise in:

Generative AI → RAG → LLM Applications → Agents → Advanced Deep Learning → MLOps → AI Engineering

Business Problem
      ↓
Data Collection & SQL Analysis
      ↓
EDA & Feature Engineering
      ↓
Machine Learning / Deep Learning
      ↓
Evaluation & Explainability
      ↓
Modular Inference Pipeline
      ↓
Application Development
      ↓
Testing & Reproducibility
      ↓
Deployment

I enjoy building projects that demonstrate not only model performance, but also the engineering required to turn models into usable applications.

🌐 Portfolio   |   💼 LinkedIn   |   📄 Resume   |   🐙 GitHub