MMCPMarkets
BZ

bmn-zyb/ai_intervieweragent

Developer tools

This page was translated automatically. Technical details may differ from the original. View the original

36 stars 2 forks Quality 51 Trend 16
Repository Install

An engineering-grade multi-agent collaboration system built on LangChain/LangGraph + LlamaIndex, supporting fully automated interview preparation, mock interviews, real-time evaluation, and review planning after uploading resumes and job JDs; The system includes multi-agent DAG orchestration, RAG multi-channel recall, MCP consensus integration, dual-engine memory system, and dynamic difficulty adjustment. It is deployed on Linux servers via Python web services, supporting multimodal interaction via camera, voice input, and text.

Overview

Upload your resume + enter your position. JD → AI will automatically conduct a complete mock interview, → score → report → develop a review plan InterviewAgent is a leading AI Agent application that integrates the most mainstream technical paradigms in the current AI engineering field: - 🤖 : 8 dedicated agents, LangGraph orchestration - 🔍 : dual vector + BM25, not just simple vector library adjustment - 🧠 : Redis short-term + MySQL long-term, remembers your weaknesses across sessions - 🎚️ : Level 3 difficulty state machine simulation, simulates real interviewer strategies - 🧩 : Stateful multi-turn interaction module, pluggable and scalable - 🔌 : Standard protocol access to external tools - 🎬 : Web interface + camera + STT + TTS voice interaction Through framework orchestration DAG collaboration, each Agent has its own independent prompt strategy and toolset. - : Semantic vector retrieval + keyword reverse search for complementary coverage - : Reciprocal Rank Fusion algorithm fuses two sources of results - : Large models reorder candidate results to ensure final quality - Manages the entire retrieval pipeline 📄 Core file: rag/retrievers/hybrid_retriever.py - : Short-term memory, manages current conversational context, O(1) Read/write - : Long-term memory, persists user profiles and historical data, supports complex queries - : Unified data access layer LangGraph StateGraph orchestration, where nodes pass data by sharing State

Overview (machine translated)

🎯 InterviewAgent · AI mock interviewer

A Comprehensive Engineering AI Agent Application

Upload your resume + enter your position. JD → AI will automatically conduct a complete mock interview, → score → report → develop a review plan

📖 Table of Contents

  • Project Overview
  • Eight core technological highlights
  • Three-tier architecture
  • Interview Process (DAG).
  • Project structure
  • Environmental preparation
  • Quick start
  • API documentation
  • Core technology stack
  • License

🌟 Project Overview

InterviewAgent is a complete engineering-grade AI Agent application that integrates the most mainstream technical paradigms in the current AI engineering field:

  • 🤖 Multi-Agent DAG Collaboration: 8 dedicated agents orchestrated by LangGraph
  • 🔍 RAG Multiplex Recall + LLM Rerank: Vector + BM25 dual channeling, not a simple vector tuning library
  • 🧠 Agent Memory System: Redis short-term + MySQL long-term, remembers your weaknesses across sessions
  • 🎚️ Dynamic Difficulty Adjustment: Three-level difficulty state machine simulates real interviewer strategies
  • 🧩 Skill System: Stateful multi-turn interaction module, expandable and pluggable
  • 🔌 MCP protocol integration: Standard protocol access to external tools
  • 🎬 Complete Project Delivery: Web interface + camera + STT + TTS voice interaction

✨ Eight Core Technology Highlights

🤖 1 · Multi-Agent DAG collaborative orchestration

It’s not a single prompt that can go, but 8 Agents each playing their role:

| Agent | Responsibilities | |-------|------| | IntentRouter | Intent Recognition and Routing Distribution | | JDAnalyzer | Parse the JD, extract the tech stack and job level requirements | | ResumeAnalyzer | Resume matching analysis to identify candidates’ strengths and weaknesses | | QuestionPlanner | Topic distribution based on JD+ resume planning | | Interviewer | Hosted multiple rounds of interviews, dynamically asking in-depth questions | | Evaluator | Score each question and generate an evaluation report | | StudyPlanner | Create a personalized review plan | | ChatAgent | Handling small talk and guided dialogue |

DAG collaboration is orchestrated through the LangChain + LangGraph framework, with each Agent having its own Prompt strategy and toolkit.

📄 > Core file: orchestration/graph.py

🔍 2 · RAG multiplex recall + LLM rerank

  • Dual Parallel: Semantic vector retrieval + keyword reversed search, complementary coverage

This page was translated automatically. Technical details may differ from the original.

Install

This server does not publish a one-line install command.

Open the repository installation guide New to MCP? Start with the mcpskill.org guide →