Dinu Yadav

About

Computer Science student at GD Goenka University specializing in AI & ML, with hands-on project experience in RAG systems, churn analytics, and NLP classification. Strong foundation in Python, machine learning, data analysis, and software development concepts.

Education

• 2026 GD
  Bachelor of Technology

  GD Goenka University

  Computer Science & Engineering (AI/ML Specialization) · 2022 – 2026

  Computer Science & Engineering (AI/ML Specialization) · 2022 – 2026
• 2022 SD
  Senior Secondary Certificate

  SDMS

  Physics, Chemistry, Mathematics · 2022

  Physics, Chemistry, Mathematics

Skills

Tools / apps / platforms

Projects

• NLP Sentiment Classification & Review Mining System
  Python, TensorFlow, NLTK, Scikit-Learn

  Designed a multi-class NLP text classification workflow to categorize user reviews into positive, negative, and neutral sentiment categories. Preprocessed unstructured text data utilizing tokenization, stop-word removal, lemmatization, and TF-IDF feature extraction. Evaluated model performance using classification matrices, precision-recall metrics, and confusion matrices to ensure balanced class predictions.

• Customer Churn Analytics & Predictive Risk Pipeline
  Python, Pandas, Scikit-Learn, Seaborn, SQL

  Analyzed multi-dimensional customer behavioral data across 7,000+ records to identify primary churn determinants and usage drop-offs. Engineered structured data pipelines with missing value imputation, categorical encoding, and correlation matrix analysis. Identified key risk cohorts through exploratory statistical analysis, recommending targeted retention interventions with an estimated 12% reduction in churn.

• Retrieval-Augmented Generation (RAG) Document Query Engine
  Python, Vector Search, LLM Integration, Embeddings

  Architected an end-to-end RAG question-answering pipeline allowing semantic search and grounded synthesis across unstructured documents. Implemented document chunking, text vectorization, and dense semantic retrieval, mitigating hallucinations and grounding answers in retrieved context. Optimized context window allocation and prompt construction to provide factual, source-attributed responses with low query latency.

Courses & certifications

• Principles of Management (Project Life Cycle, Resource Planning & Risk Management) · NPTEL / IIT Roorkee
• Computer Vision (Foundations of Image Processing, Feature Extraction & Visual Recognition) · NPTEL / IIT Kharagpur