Jeremy Burke
Project 03
Contents
AI · Business

Guest Intelligence Dashboard

A five-day onsite sprint that merged 44,874 guest records from four disconnected systems into a single source of truth, with a dashboard that shows operators who their guests are, what they prefer, and when they return as regular or VIP guests.

Summary

01

Worked onsite with the client during a five-day sprint.

02

Merged 44,874 guest records from four disconnected systems into a single source-of-truth database.

03

Built a data dashboard to visualize and understand guest profiles.

Role
Solo operator embedded with the client.Build, data audit, and client delivery.
For
The Log CabinMeridian XYZ
When
Five-day sprint, April 2026
Built with
Next.js, React, TypeScript, Recharts, D3, Python, pandas, RapidFuzz, SQLite, Anthropic API

Description

Context

A historic fine-dining restaurant had guest records spread across OpenTable, Toast, and Mailchimp. Because the systems were disconnected, the team could not easily distinguish regulars from first-time visitors.

Approach

I spent five days onsite learning how the team used each system and where the records overlapped. I then built a pipeline that matched and merged guest profiles using phone numbers, email addresses, and name similarities.

Deliverables

The team received a guest intelligence dashboard running on its own infrastructure. It brings together guest profiles, visit history, repeat behavior, VIP scoring, and email-list health, giving staff better information for daily restaurant operations.

More projects