How itsbk13/ai-interview-copilot-repo works
Interview Copilot is a Streamlit recruiting tool for loading candidates, conducting AI-assisted interviews, and reviewing candidate records. The main workflow routes an interviewer from candidate setup into an interview workspace, where an LLM can request registered MCP tools; those tools read and update candidate data. Separate pages support resume uploads and candidate search. The graph includes all indexed core modules. Some page-to-tool links rely on the SOURCE INDEX’s import references; unsampled behavior is not inferred beyond those references.
Main entry points: app.py Page Router · models.py Data Models · mcp_server.py MCP Tool Server · interview_workspace.py Interview Workspace
Components 20
Interview UI
- Page Router
app.py - Navigation Sidebar
frontend/components/sidebar.py - Home Dashboard
frontend/pages/home.py - New Interview
frontend/pages/new_interview.py - Interview Workspace
frontend/pages/interview_workspace.py - Candidate Browser
frontend/pages/candidate_db.py - Resume Upload
frontend/pages/upload_resume.py - UI Components
frontend/components/ui_components.py
Interview Services
- LLM Tool Agent
frontend/llm_client.py - MCP Tool Server
server/mcp_server.py - Resume Processing
tools/resume_tools.py
Candidate Records
- Candidate Reads
tools/read_tools.py - Candidate Search
tools/query_tools.py - Interview Writes
tools/write_tools.py
Data Storage
- Seed Data
database/seed_data.py - Data Models
database/models.py - SQLite Database
database/interview_copilot.db - Resume PDFs
resumes
Other components
- Interviewer
- Groq API
Connections 33
- Interviewer → Page Router: opens
app.py - Page Router → Navigation Sidebar: renders
app.py - Page Router → Home Dashboard: routes to
app.py - Page Router → New Interview: routes to
app.py - Page Router → Candidate Browser: routes to
app.py - Page Router → Resume Upload: routes to
app.py - Page Router → Interview Workspace: routes to
app.py - New Interview → Interview Workspace: starts interview
- New Interview → Resume Processing: uses
frontend/pages/new_interview.py - New Interview → UI Components: renders with
frontend/pages/new_interview.py - Interview Workspace → LLM Tool Agent: requests agent
frontend/pages/interview_workspace.py - Interview Workspace → Candidate Reads: gets questions
frontend/pages/interview_workspace.py - Interview Workspace → Interview Writes: saves results
frontend/pages/interview_workspace.py - Interview Workspace → UI Components: renders with
frontend/pages/interview_workspace.py - Home Dashboard → Candidate Reads: reads records
frontend/pages/home.py - Candidate Browser → Candidate Reads: loads candidates
frontend/pages/candidate_db.py - Candidate Browser → Candidate Search: filters candidates
frontend/pages/candidate_db.py - Candidate Browser → UI Components: renders with
frontend/pages/candidate_db.py - Resume Upload → Resume PDFs: writes PDF
frontend/pages/upload_resume.py - LLM Tool Agent → MCP Tool Server: discovers and executes
frontend/llm_client.py - LLM Tool Agent → Groq API: sends tool requests
frontend/llm_client.py - MCP Tool Server → Resume Processing: registers tools
server/mcp_server.py - MCP Tool Server → Candidate Reads: registers tools
server/mcp_server.py - MCP Tool Server → Candidate Search: registers tools
server/mcp_server.py - MCP Tool Server → Interview Writes: registers tools
server/mcp_server.py - MCP Tool Server → Data Models: initializes database
server/mcp_server.py - Resume Processing → Data Models: reads and writes
tools/resume_tools.py - Resume Processing → Resume PDFs: loads PDF
tools/resume_tools.py - Candidate Reads → Data Models: queries
tools/read_tools.py - Candidate Search → Data Models: queries
tools/query_tools.py - Interview Writes → Data Models: updates
tools/write_tools.py - Data Models → SQLite Database: persists records
database/models.py - Seed Data → Data Models: uses models
database/seed_data.py