Supa - AI Platform for Event Management
🎉 Supa - AI Platform for Event Management (UK)
Workflow-driven system for event process management and automation
📌 Context and Challenge
The Supa project addresses a classic problem in the event industry:
- communication with suppliers is manual (email, messengers)
- contractor selection is chaotic
- date and price negotiation is a lengthy process
- information is fragmented (spreadsheets, emails, calendars)
Engineering task:
to consolidate all event organization processes into a single manageable system with AI automation
🎯 System Goal
- automate interaction between participants
- accelerate supplier selection and terms negotiation
- reduce manual correspondence
- make the process transparent, trackable, and reproducible
🧠 Architectural Approach
Core Principle:
AI-first + workflow orchestration + multi-tenant platform
The system is built as:
- a set of domain services
- a unified AI assistant layer
- an event-driven interaction model
🧩 Key Modules
👤 1. User Accounts (Multi-role Accounts)
Support for various roles:
- event organizers
- suppliers (catering, venues, logistics)
- managers
Functionality:
- role-based access
- data isolation (multi-tenant)
- custom scenarios per role
✉️ 2. Communication Module (Messaging Engine)
Internal messaging system between accounts:
- organizer ↔ supplier dialogues
- communication history storage
- message status
🤖 AI Email Generation
Key feature:
-
email generation based on request templates:
-
date inquiry
-
price inquiry
-
service clarification
-
adaptation to supplier context
Technically:
- LLM + prompt templates
- context = dialogue history + supplier profile
🧠 3. Manager AI Assistant (Core AI Layer)
Central component of the system.
Functions:
📅 Date Availability Check
- supplier calendar analysis
- consideration of business rules (buffer, load)
📌 Booking via Chat
-
user types: "book for June 15th"
-
system:
-
validates slot
-
creates event
-
records booking
🍽 Menu and Price Selection
-
parsing supplier documentation
-
extracting:
-
menu
-
prices
-
recommendations based on budget
🔍 Supplier Selection
-
search by database
-
filtering:
-
price
-
service type
-
availability
⚙️ How it's Implemented
AI Assistant = orchestration layer:
- intent detection
- request routing
- calling domain services
📬 4. RFQ (Request for Quotation) Engine
Automation of quotation requests:
- request is formed
- sent to suppliers
- responses are aggregated
- best option is selected
📦 5. Supplier Booking
- recording the selected contractor
- date blocking
- link to event
📅 6. Event Calendar
Functionality:
- create / edit events
- display participants
- synchronization with bookings
Peculiarity:
calendar = source of truth for the entire system
🚚 7. Logistics Calculation
- delivery cost calculation
- distance consideration
- integration with supplier rates
🏗 System Architecture
🧱 Microservice Structure
- Account Service
- Messaging Service
- AI Service
- Supplier Service
- Booking Service
- Calendar Service
- Pricing/Logistics Service
🔄 Event-Driven Model
-
events:
-
message_sent -
supplier_selected -
booking_created -
event_updated
Used for:
- service synchronization
- reactive logic
🧠 AI Layer
Separate service:
- LLM orchestration
- prompt management
- tool calling
⚙️ Process Orchestration
Approach analogous to fintech:
each event = workflow
Example:
Create Event →
Select Suppliers →
Request Quotes →
Compare →
Book →
Manage Event →
Close