Shyam Gadhiya
Artificial Intelligence and Machine Learning Engineering B.E
Experience
AI – ML Intern
Ascendance Solutions
Developed and deployed ML and GenAI solutions using Python, improving task automation efficiency by 30–40%. Designed and deployed an end-to-end automated supply chain monitoring system using n8n, integrating Agentic AI (ARIA) to autonomously analyze root causes for shipment delays. Implemented workflow logic that triggers alternative route evaluation and auto-generates executive reports, improving task automation efficiency by 30-40%.
Summer Internship Program – Data & Ai + Azure Cloud
Microsoft
Completed 60 hours of hands-on training under Microsoft & GTU initiative. Gained practical skills in Azure AI, Natural Language Processing, Generative AI on Azure ML, and Power BI. Built NLP and Deployed solutions and dashboards using Azure ML, Power BI, and Cognitive Services.
Education
L.D. College Of Engineering
B.E
CPI/7.77
G.G. Zadafiya Vidhyalaya
HSC
82
G.G. Zadafiya Vidhyalaya
SSC
84.05
Skills
PROJECTS
Multi Agent Research Assistant
Architected an autonomous multi-agent research system using LangGraph and Gemini, implementing a Corrective RAG (C-RAG) framework to verify retrieval quality and reduce hallucinations by ~30%. Orchestrated a dual-retrieval pipeline using Tavily for real-time web intelligence and FAISS for local document synthesis, ensuring citation-backed accuracy for high-stakes reports. Developed a real-time execution streaming UI via Streamlit, providing transparency into the agent's task decomposition and reasoning process.
Pneumonia Detection using CNN
Developed a Convolutional Neural Network (CNN) to automatically detect pneumonia from chest X-ray images. Trained the model on a large labeled dataset, enabling it to distinguish between normal and pneumonia lung. Achieved a test accuracy of ~92%, demonstrating strong potential for real-world clinical diagnostic support.
E-commerce Product Recommendation Chatbot using Generative AI (RAG Framework)
Built an intelligent recommendation system that assists users in finding electronics products based on preferences like price, rating, and durability. Integrated Retrieval-Augmented Generation (RAG) architecture using LangChain, FAISS, and Hugging Face Transformers (DistilGPT-2). Processed and analyzed the Amazon Electronics Reviews dataset (Kaggle) — extracted key insights, embedded reviews using SentenceTransformers, and created a vector database for semantic search.