customer_support · workflow
Agoda builds a multi-agent Booking Form Bot to answer traveler questions at checkout
Users frequently abandoned the booking form due to unanswered questions about promotions, cancellation rules, and pricing; 28% returned to the property page to re-check details, and Payment & Promo Terms had only a 4% click-through rate despite spanning over 50 mobile pages.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · User opens chat on booking form
Users tap a chat icon on the booking form to open a conversational interface and ask questions in natural language.
Tools used
OpenAIProperty API (PAPI)
Outcome
The Booking Form Bot answers 89% of user queries fully or partially, delivering real-time context-aware answers without requiring users to leave the booking page.
Results
Volume28%
Source
https://medium.com/agoda-engineering/building-agodas-booking-form-bot-the-agentic-way-407455e50808
Grounding & classification
Source type: technical build writeup
24 fields verified against source quotes.
agentic workflowchatbotconversational aimulti agent workflowchat transcriptknowledge basebuilder submittedmetric backednamed customerproduction runtime claimedtools describedworkflow describedtravelcustomer satisfactiondeflection ratetechnical build writeupcustomer supportecommerce opsagentic task executionextract classify route