How guptaastha1427/trip-planner works
This is a Streamlit trip-planning app with a separate demo travel portal and policy Q&A. The main user flow collects trip preferences, validates them, runs an itinerary agent using map data and Gemini, then displays and saves the result; users can also view maps, get trip details, and give feedback. Separate branches cover bookings, refunds, support, and policy search, with evaluation tooling for retrieval and answer quality. Some links reflect README documentation or SOURCE INDEX references rather than inspected call sites; those limits are noted in the evidence fields.
Main entry points: app.py Planner App · agent.py Itinerary Agent · bookings_refunds.py Bookings & Refunds · poi_search.py POI Search
Components 28
Trip Planner
- Planner App
app.py - Itinerary Agent
services/trip/agent.py - POI Search
services/trip/poi_search.py - Trip History
services/trip/trip_history.py - Trip Feedback
services/trip/feedback.py - Map View
services/trip/map_viz.py - Geocoding
services/trip/geocoding.py
Policy Q&A
- Policy Pages
pages/portal/terms_policies.py - Policy RAG
services/rag/website_rag.py - Support Answers
services/rag/support_faq.py - Assistant UI
services/ui/rag_chat_ui.py
Travel Portal
- Support Page
pages/portal/support.py - Assistant Auth
services/portal/rag_chat_auth.py - Bookings & Refunds
pages/portal/bookings_refunds.py - Booking & Refund Logic
services/portal/bookings.py - Refund Eligibility
services/portal/refund_eligibility.py - Portal State
services/portal/customer_portal_store.py - Account & Admin
pages/portal/account_status.py - Portal User Context
services/portal/portal_user_context.py
Shared Services
- Input Validation
services/core/validation.py - HTTP Client
services/core/http_client.py
Evaluation
- RAG Evaluation
eval/runners/ragas_pipeline.py - Eval Embeddings
eval/helpers/real_embeddings.py - Eval Retrieval
eval/helpers/terms_retrieval.py - Eval Orchestrator
eval/run_pipeline.py
Other components
- Traveler
- OpenStreetMap
- Gemini
Connections 31
- Traveler → Planner App: enters
README.md - Planner App → Input Validation: validates inputs
app.py - Planner App → Itinerary Agent: runs itinerary
app.py - Itinerary Agent → POI Search: searches places
services/trip/agent.py - Itinerary Agent → Input Validation: validates places
services/trip/agent.py - POI Search → Geocoding: resolves location
services/trip/poi_search.py - Geocoding → HTTP Client: requests data
services/trip/geocoding.py - POI Search → HTTP Client: requests places
services/trip/poi_search.py - HTTP Client → OpenStreetMap: queries
services/core/http_client.py - Itinerary Agent → Gemini: generates itinerary
README.md - Planner App → Trip History: saves trips
app.py - Planner App → Trip Feedback: records feedback
README.md - Planner App → Map View: renders map
app.py - Traveler → Policy Pages: opens policies
README.md - Policy Pages → Policy RAG: searches policies
- Traveler → Support Page: opens support
README.md - Support Page → Support Answers: answers questions
- Support Page → Assistant UI: offers assistant
- Assistant UI → Assistant Auth: checks session
services/ui/rag_chat_ui.py - Policy RAG → Gemini: embeds and answers
README.md - Traveler → Bookings & Refunds: manages bookings
README.md - Bookings & Refunds → Booking & Refund Logic: loads booking logic
pages/portal/bookings_refunds.py - Bookings & Refunds → Refund Eligibility: checks eligibility
pages/portal/bookings_refunds.py - Bookings & Refunds → Portal State: reads and writes
pages/portal/bookings_refunds.py - Traveler → Account & Admin: views account
README.md - Account & Admin → Portal User Context: resolves customer
- Portal User Context → Portal State: reads profiles
services/portal/portal_user_context.py - Eval Orchestrator → RAG Evaluation: runs evaluations
eval/run_pipeline.py - RAG Evaluation → Eval Embeddings: builds eval index
eval/runners/ragas_pipeline.py - RAG Evaluation → Eval Retrieval: measures retrieval
eval/runners/ragas_pipeline.py - Eval Retrieval → Policy RAG: uses retrieval logic
eval/helpers/terms_retrieval.py