adityajakkula7/askiiita

How adityajakkula7/askiiita works

AskIIITA is a campus-information RAG chatbot with a Streamlit interface and Flask API endpoints. Users sign in, ask questions, and receive answers grounded in retrieved institutional documents; admins can inspect query audit logs. Separately, ingestion loads and chunks PDF/TXT documents, adds access levels, embeds them, and stores them for retrieval. Authentication and role-based access span the chat and API paths. The README documents the overall RAG stack; app.py excerpts are partial, so some chat-side service links rely on that documentation rather than inspected call sites.

Main entry points: query.py Query Endpoint · auth.py Auth Helpers · app.py Chat RAG · audit_logs.py Audit Endpoint

Components 17

Chat Experience

Streamlit Chat app.py
Chat RAG app.py

API Services

Login Endpoint api/login.py
Registration Endpoint api/register.py
Query Endpoint api/query.py
Audit Endpoint api/audit_logs.py

Knowledge Pipeline

Campus Documents data
Document Ingestion ingest.py
Local Vector Search local_vector.py
Atlas Vector Store
Gemini Embeddings
Gemini LLM

Identity and Audit

Auth Helpers auth.py
Audit Records api/query.py
User Records auth.py

Other components

User
Admin

Connections 25

User → Streamlit Chat: asks questions README.md
Streamlit Chat → Chat RAG: submits queries app.py
Streamlit Chat → Auth Helpers: authenticates app.py
Chat RAG → Atlas Vector Store: retrieves chunks README.md
Chat RAG → Gemini Embeddings: embeds query README.md
Chat RAG → Gemini LLM: generates answer README.md
User → Login Endpoint: signs in api/login.py
User → Registration Endpoint: registers account api/register.py
User → Query Endpoint: submits query api/query.py
Admin → Audit Endpoint: requests logs api/audit_logs.py
Login Endpoint → Auth Helpers: verifies credentials api/login.py
Registration Endpoint → Auth Helpers: creates user api/register.py
Query Endpoint → Auth Helpers: verifies token api/query.py
Audit Endpoint → Auth Helpers: verifies token api/audit_logs.py
Auth Helpers → User Records: reads and writes auth.py
Campus Documents → Document Ingestion: loads files ingest.py
Document Ingestion → Gemini Embeddings: embeds chunks ingest.py
Document Ingestion → Atlas Vector Store: stores chunks ingest.py
Query Endpoint → Atlas Vector Store: retrieves chunks api/query.py
Query Endpoint → Local Vector Search: falls back to search api/query.py
Local Vector Search → Gemini Embeddings: embeds query local_vector.py
Local Vector Search → Campus Documents: loads source files local_vector.py
Query Endpoint → Gemini LLM: generates answer api/query.py
Query Endpoint → Audit Records: writes query log api/query.py
Audit Endpoint → Audit Records: reads recent logs api/audit_logs.py

More HTML diagrams: chenjackle45/html-visualizer · deptofdefense/dds-at-defcon · skanga/dbchat · thedefibat/road-accidents-prediction-and-classification · allenai/break

Drawn by GitDiagram on October 1, 2026 from the repository's file tree, README and source files. Read as Markdown ·