TERA — Event Revenue Agent
Seven AI agents that build a trade show's sponsorship pipeline — and stop for a human "yes" at every consequential step.
For B2B trade shows and confexes: agents find potential sponsors, score them, identify the right contact, write briefings, draft outreach, and build renewal cases. Three human gates stand between the AI and the outside world — nothing contacts a company, and nothing sends, without a person deciding it should.
Sponsorship sales is research-heavy, repetitive, and permanently behind — exactly the shape of work agents are good at and exactly the domain where an unsupervised agent could do real damage. TERA is built around that tension. Seven specialized agents run the pipeline: discover companies, score fit, find the person, brief the seller, draft the outreach, build the renewal case. Three human gates interrupt it: before a company enters the pipeline, before any contact, before any send. Multi-tenant from the foundation (each event organization is a tenant with row-level isolation), with a real event as its design partner and live runs already proving the discovery and matching mechanics — one re-run took an exhibitor extraction from nothing to a full list. A read-only demo has run for its design partner since September.
Stack
- .NET 8
- Azure Service Bus
- Azure SQL
- Azure Container Apps
- React
- Clerk
- Anthropic Claude
Have a revenue process that's really a research process?
Let’s talk