How vishal060803/ai-powered-itsm-helpdesk-automation works
This is a prototype IT helpdesk portal: employees ask support questions or search approved guidance, receive grounded troubleshooting, and can use a simulated ticket/request dashboard. The implemented runtime centers on the React portal, FastAPI knowledge routes, and a document-to-vector retrieval pipeline. MongoDB is initialized as a persistence foundation, but README describes ticket persistence, ServiceNow integration, self-heal automation, and software provisioning as future work. Edges reflect shown calls or README-documented flows; unsampled UI and backend behavior is not inferred beyond that documentation.
Main entry points: knowledge.py Knowledge routes · App.jsx Portal UI · main.py FastAPI app · knowledge_pipeline.py Chunk & embed
Components 15
Employee portal
- Portal UI
frontend/src/App.jsx - App mount
frontend/src/main.jsx - Dashboard & requests
frontend/src/App.jsx
Backend API
- FastAPI app
backend/app/main.py - Health routes
backend/app/routers/health.py - Knowledge routes
backend/app/routers/knowledge.py - API contracts
backend/app/schemas.py
Knowledge retrieval
- Grounded assistant
backend/app/chat_assistant.py - Vector search
backend/app/knowledge_vector_store.py - Chunk & embed
backend/app/knowledge_pipeline.py - Approved articles
data/knowledge_docs - Embedding chunks
data/embeddings/knowledge_chunks.jsonl
Persistence foundation
- Mongo setup
backend/app/database.py
Other components
- Employee
- MongoDB
Connections 15
- Employee → Portal UI: uses
README.md - App mount → Portal UI: renders
frontend/src/main.jsx - Portal UI → Knowledge routes: requests search
frontend/src/App.jsx - Portal UI → Knowledge routes: asks questions
README.md - Portal UI → Dashboard & requests: shows
README.md - FastAPI app → Health routes: registers
backend/app/main.py - FastAPI app → Knowledge routes: registers
backend/app/main.py - FastAPI app → Mongo setup: initializes
backend/app/main.py - Mongo setup → MongoDB: connects
backend/app/database.py - Knowledge routes → Vector search: searches
backend/app/routers/knowledge.py - Knowledge routes → Grounded assistant: builds reply
backend/app/routers/knowledge.py - Knowledge routes → API contracts: uses models
backend/app/routers/knowledge.py - Approved articles → Chunk & embed: provides articles
README.md - Chunk & embed → Embedding chunks: writes chunks
README.md - Embedding chunks → Vector search: indexes chunks
backend/app/knowledge_vector_store.py