The Future of Registration: RunSignup Demos Buy in Chat

Why Buy in Chat Will be a Difference Maker for Events

Buy in chat might sound duplicative – simply another way to accomplish a registration or ticket purchase, when online registration already works. But there are a few key differentiators that will make buy in chat beneficial for races and other events:

• Simplified Transactions: Consumers are always looking for ways to make purchasing easier, especially as transactions become more common on mobile devices. It’s the reason that at least 30% of buyers opt for Apple Pay when it’s an option. Over time, more automation will occur, with LLMs automatically knowing key attendee data, and purchases via chat will become easier and more efficient.
• Research + Buy: RunSignup is working to add our MCP tools to major LLMs, giving them access to all our public information about races and other events. This means that potential event attendees will be able to research upcoming events in an LLM like ChatGPT – with TicketSignup’s races easily accessed by the LLM through the TicketSignup’s MCP. Then, once a person identifies the race they are interested in, they will be able to complete the transaction immediately, all without leaving their ChatGPT conversation.

In short: finding events and buying tickets will be easier, with fewer opportunities for potential attendees to abandon the event.

Buy In Chat in Action

Yesterday, TicketSignup developer Duy Nguyen demoed a prototype showing the coming functionality of Agentic Commerce using a simple Ticket Event. As you can see, the AI purchase is able to correctly select the right tickets and quantity, request the required information, and parse information given in natural language to identify what needs to be recorded in your TicketSignup report (such as last name).

See “Buy-in-Chat” in action:

Is Buy in Chat Secure?

Yes, buy in chat will be secure. RunSignup’s approach ensures security for sensitive data like personal identifiers and credit card numbers. To do this, RunSignup’s servers sanitize sensitive information like credit card information and swap it out for a placeholder string before sending it to an LLM. Essentially, RunSignup takes advantage of the LLM to facilitate natural language conversation, while keeping all data protected within the RunSignup environment.

Watch Duy demo the security setup:

When Can We Buy in Chat?

Currently, this functionality is a prototype we are testing and is not yet available to customers, even for ticket events. While we don’t yet have a date for release, there are a few things happening in the backend to make buy in chat a reality in the near future:

• We are continuing to build out the RunSignup and TicketSignup chat infrastructure for our own agentic chatbot.
• Adyen, the backend Payment processor for RunSignup, is refining their MCP tools so LLMs can access it to complete transactions.
• Adyen is working with major LLMs like ChatGPT and Google to understand and meet the standards of each LLM for transaction processing to facilitate processing within those interfaces.

For a more complete look at the AI tools that RunSignup is working with to create a better future for events, check out our full webinar on Open API and MCP.

Summaries of the Content of Included Videos

For anyone unable to watch the full demos, a summary of the included content of each can be found below.

Summary of Agentic Transaction Process Demo

• The team is developing agentic transaction flows that allow users to complete registrations and purchases directly through a chat interface.
• Instead of navigating forms, users can stay on the event page and interact conversationally with an AI assistant.
• Natural language ticket selection:
  • Users can request tickets using everyday language (e.g., “I’d like to buy two of the cheapest tickets”).
  • The system identifies the appropriate ticket type and quantity.
  • Users can update their request naturally (e.g., “Actually, I want three”), and the cart updates accordingly.
• Registration details:
  • The AI collects basic information like first name, last name, and email.
  • It can extract the correct data even when surrounded by extra natural-language text (e.g., “My last name is Nguyen”).
• Payment handling:
  • Users can provide all payment details in a single natural-language message.
  • The system parses this into the required structured fields (credit card number, expiration, CVV, etc.).
  • Additional fields like ZIP code and country can also be provided conversationally.
• Waiver agreement:
  • The assistant displays the standard event waiver.
  • Users can accept it with a simple natural-language confirmation (e.g., “I want to sign off on this waiver”).
• End-to-end flow:
  • With all steps completed, the system finishes the transaction entirely through chat.
  • Demonstration concludes with successful purchase of three tickets.

Summary of Security Approach

• The system is designed to keep all payment and personal information secure throughout the process.
• When a user sends sensitive data (e.g., credit card information), it is first handled entirely on the RunSignup side.
• Sanitization process:
  • The server inspects the incoming prompt.
  • Sensitive information is identified.
  • All sensitive values are replaced with placeholder strings.
  • Only the sanitized version is sent to the LLM.
• LLM involvement:
  • The LLM (e.g., OpenAI) receives only sanitized data.
  • It interprets the prompt and determines the next tool or action—such as calling the “process payment” tool.
  • The placeholder strings are used as inputs during this step.
• Repopulation of sensitive data:
  • When the request returns to the RunSignup system, cached sensitive data is securely reinserted into the process.
  • This ensures accurate processing without exposing private information.
• Key guarantee:
  • At no point does sensitive or personal data leave RunSignup’s servers.
  • Natural language processing benefits from the LLM, but all sensitive details remain protected on RunSignup’s side.