Hello, I'm Deepak Kumar 👋🏻

A machine learning pipeline that forecasts future sales based on historical data.

• ●Analyzed exploratory data using Pandas and Matplotlib
• ●Trained a regression model with Scikit-Learn to predict future outcomes
• ●Deployed the model inference via a REST API

A recurrent neural network based chatbot that processes text and infers intent.

• ●Designed a custom NLP pipeline tokenizing user inputs
• ●Trained embeddings and a Seq2Seq model using TensorFlow
• ●Implemented an interactive feedback loop to tune answers

A real-time object detection and image classification system.

• ●Utilized OpenCV for video stream capture and basic filtering
• ●Applied a pre-trained ResNet model in PyTorch for high accuracy detection
• ●Optimized inference time for real-time video processing

A K-Means clustering application grouped customers into market segments.

• ●Processed raw transaction data into feature vectors
• ●Determined optimal clusters using Silhouette score analysis
• ●Visualized demographic bounds and clusters using Seaborn