BlockShop: Towards a Decentralized and Intelligent Online Marketplace
Surya Pratap1, Suraj Yadav2, Nitesh Nisad3, Ankit Singh4
Department of Computer Science and Engineering, Prasad Institute of Technology, Jaunpur, Uttar Pradesh, India
-Guided By: Ayush Yadav
Abstract - Traditional automated interview systems often rely on static question banks, resulting in a generic and impersonal candidate experience. Such systems frequently fail to accurately assess an applicant's unique skill set. VeriHire addresses these challenges by introducing an adaptive, AI-powered interview platform designed to conduct dynamic and personalized interviews.
The system begins by analysing a candidate’s professional profile — including resumes, portfolios, and repositories — to understand their expertise. Using Natural Language Processing (NLP) and a Large Language Model (LLM), VeriHire generates tailored, real-time questionnaires and adapts subsequent questions based on the candidate’s live responses. This approach enables deeper exploration of technical knowledge, problem-solving skills, and project experience, closely simulating the process of a human interviewer.
MCP (Model Context Protocol) is a new standard that allows language models to securely communicate with external tools, databases, and APIs, making AI systems more extensible and capable. RAG (Retrieval-Augmented Generation) is a powerful technique that improves model accuracy by fetching relevant information from a knowledge source before generating a response. LangChain is a popular open-source framework that helps developers build applications using language models by chaining prompts, tools, and memory into a structured workflow
The system begins by analysing a candidate’s professional profile — including resumes, portfolios, and repositories — to understand their expertise. Using Natural Language Processing (NLP) and a Large Language Model (LLM), VeriHire generates tailored, real-time questionnaires and adapts subsequent questions based on the candidate’s live responses.
Key Words: MCP, RAG, LANG-CHAIN
