Nivesh Tyagi

Machine Learning Intern with practical experience in algorithm development, predictive modeling, and data mining. Proficient in statistical programming languages such as Python and R, with a solid grasp of machine learning principles. Demonstrated success in utilizing big data analysis to enhance project outcomes and operational efficiency. Dedicated to continuous learning in artificial intelligence, deep learning, and data architecture.

• Participated in national and inter-college hackathons, building ML and automation-based solutions.
• Member of tech and AI/ML communities, contributing to coding events and peer knowledge sharing.
• Authored technical blogs on Python, machine learning, and project development.
• Volunteered in tech education and social impact initiatives, promoting digital literacy.
• Pursued interests in UI/UX design and frontend development for personal projects.
• Completed several online certifications and participated in Kaggle competitions to enhance practical skills.

• AI Doctor (LLM + Voice & Vision): Developed a multimodal AI health assistant using Gemini 1.5 Pro, RAG, LangChain, and speech/image input to deliver real-time diagnosis and treatment suggestions.
• AI-Powered Blog Generator: Built an automated blog creation platform using LLaMA 3.1 8B, AWS Bedrock, RAG, and Streamlit, improving enterprise content workflows.
• Conversational AI Chatbot: Designed a customer support bot with GPT-4o, LangChain, and Bedrock, achieving a 40% reduction in query resolution time via intelligent response handling.
• AutoML Studio: Created a no-code ML platform using MLflow, SHAP, and FastAPI for building classification, regression, and time-series models with data drift detection and explainability.
• DocuWiz AI: Engineered a document analysis tool with BART, DistilBERT, and RAG for PDF/DOCX summarization and sentiment analysis; reduced review time by 60%.
• Harvestify - AI Plant Health Assistant: Deployed a DL-based crop advisor using Azure AI Studio, achieving 70% faster disease detection and improving crop yield prediction by 30%.
• Document Summarization Tool: Built a summarizer using Pegasus and FastAPI, generating concise insights from lengthy documents to boost legal/academic productivity.
• Customer Satisfaction Prediction (MLOps): Designed a prediction pipeline using ZenML, MLflow, and MLOps best practices to forecast customer satisfaction with 93% accuracy.
• Bank Management System: Developed a secure banking solution in C++, handling transactions, user auth, and reporting with object-oriented design.
• Landing Page + Personal Portfolio: Designed modern, responsive web interfaces using HTML5, CSS3, JavaScript, and deployed on GitHub Pages.
• Plant Disease Prediction System: Implemented a deep learning model using CNNs and OpenCV to detect crop diseases from images; deployed via Streamlit for real-time use.

• Exploratory Data Analysis for Machine Learning
  Authorized by IBM, Coursera
• Oracle Database Foundations
  Authorized by LearnQuest, Coursera
• Cloud Computing and Distributed Systems
  8-week certification course, NPTEL