Manish Rana

I've built machine learning, deep learning, computer vision, and NLP applications using Python, SQL, and practical data analysis workflows.

During my Data Science Internship at Sabudh Foundation, I completed end-to-end training in data collection, preprocessing, exploratory data analysis, model training, evaluation, and deployment. I developed a handwriting-to-font synthesis pipeline using image preprocessing, segmentation, OCR-assisted annotation, and font generation.

My projects include an agentic audio retrieval and editing system using WavLM/MERT embeddings, Qdrant vector search, LangGraph, MongoDB, and FastAPI. I also built a semantic movie recommendation engine with TF-IDF and cosine similarity, plus a Streamlit-based system that converts handwritten samples into installable TrueType fonts.

I'm comfortable working with structured and unstructured data, from feature engineering and model evaluation to AI-driven applications. I've also earned six Dataiku certifications spanning machine learning, generative AI, application development, data engineering workflows, and MLOps.