Zest’s new AI agent finds restaurants and books reservations over text

Zest, a startup focused on conversational AI, has launched a text‑based restaurant concierge that can read a user’s credit‑card transaction history via Plaid, recommend dining options, and book tables through OpenTable, Resy, and SevenRooms—all through SMS. The service eliminates the need for a mobile app, allowing users to reserve a table with a simple text message. Zest’s AI agent processes the user’s text, queries Plaid for recent spending patterns, and then uses the reservation APIs of the partner platforms to secure a table. The solution is positioned as a frictionless, on‑demand experience that leverages existing payment data to personalize recommendations.

For enterprise IT teams, this launch highlights several critical considerations. First, the integration of a third‑party AI agent that accesses sensitive financial data via Plaid raises data‑privacy and compliance questions. Organizations must evaluate whether their SMS gateways and data‑storage policies can accommodate the flow of credit‑card information, especially under regulations such as UK GDPR and the EU’s ePrivacy Directive. Second, the reliance on multiple reservation APIs introduces operational dependencies that require robust monitoring and rate‑limit management. Third, the use of SMS as the primary channel demands secure messaging infrastructure, end‑to‑end encryption, and compliance with telecom regulations.

Vendors involved include Plaid (payment data aggregation), OpenTable, Resy, and SevenRooms (reservation management), and the SMS platform that Zest uses (likely Twilio or similar). The broader industry implication is a shift toward conversational, data‑driven concierge services that bypass traditional mobile apps. This trend forces enterprises to re‑examine their data‑sharing agreements, API governance, and user‑experience strategies. It also signals that AI‑powered, text‑centric services will become more common, pushing vendors to offer secure, compliant API layers and for IT teams to develop new monitoring and risk‑management frameworks.

1. EVALUATE - Audit existing SMS gateway architecture: encryption, logging, and retention policies. - Review current data‑privacy framework: how credit‑card data is stored, processed, and shared. - Map API usage limits for Plaid and reservation platforms; identify potential bottlenecks. - Assess compliance posture against UK GDPR, ePrivacy, and any industry‑specific regulations.

2. PROPOSE - Quantify potential cost savings from reduced mobile app development and maintenance. - Estimate user engagement uplift: e.g., 20% increase in reservation conversions due to instant SMS booking. - Benchmark API call volumes against current limits; present projected growth. - Include ROI calculation: cost of SMS traffic vs. revenue from increased bookings.

3. TOOLS TO CONSIDER - Plaid for payment data aggregation. - Twilio or Vonage for SMS delivery and compliance. - OpenTable, Resy, SevenRooms APIs for reservation handling. - Splunk or Elastic Stack for monitoring API usage and security events. - Okta or Azure AD for identity and access management of the integration.

4. RISKS TO FLAG - Technical: API rate limits, latency, data consistency. - Compliance: Unauthorized access to credit‑card data, GDPR data‑subject rights, ePrivacy SMS consent. - Operational: Vendor lock‑in, dependency on third‑party uptime, potential data breach exposure.

5. QUICK WIN - Deploy a sandbox integration with Plaid and Twilio to test data flow and encryption. - Configure automated alerts for API failures and exceedance of rate limits. - Draft a data‑processing addendum with Plaid to clarify liability and data handling.

6. LONG-TERM PLAY - Build a secure, scalable microservice that abstracts the AI concierge, enabling future expansion to other services (e.g., travel, retail). - Implement a unified policy engine to enforce GDPR and ePrivacy controls across all conversational channels. - Conduct periodic penetration testing and third‑party audits of the SMS and API integration. - Develop a governance framework for continuous monitoring of vendor SLAs and compliance updates.