Lexis — the master database everything else depends on.
The MICE record system behind the iTravelle ecosystem: hotels, dining, attractions and airports across 139 countries. The biggest workstream of the programme, and the one that decided whether every surface above it could be trusted.
One record system, four departments
Lexis is where supplier truth lives. Every proposal, booking and invoice above it resolves back to a record here, which is why the schema mattered more than the screens.
This case study is scaffolded and waiting for your content. The structure, styling and the two Lexis assets already in the archive are in place.
- The one-paragraph summary of what Lexis is, for someone who has not read the iTravelle study
- Problem / What I did / Outcome, three short cards
- Any numbers you want to lead with beyond the three above
The problem
The iTravelle study already documents two Lexis defects you can expand here: finding a venue meant scrolling thousands of rows and filtering blind, and generic field types produced concatenated currencies and raw record IDs surfacing to operators.
- The defects, with the evidence screenshots if you have them
- What each one cost the operator in real time
Approach
How you got from the audit to the schema decision. The iTravelle study covers the argument against migration, which belongs here in more detail.
- The process steps, four or five
- The governance rule: only a data steward creates new field types
- The reference products you studied and the lesson from each
The work
Six custom field types, not generic inputs. Each replacement has its own control, validation and display.
The six field types table is already written in the iTravelle case study under The work, Lexis. Move or expand it here, with the record screens once you export them.
- Hotel record screen, commercial terms locked for Partner Ops
- The field type table: behaviour and the defect each retires
- The shared navigation, badge and record patterns
Outcome
What the schema work produced, and what you would do differently.
Got a data problem wearing an interface problem's clothes?
Most of them are. Tell us what is breaking and we will find where it actually starts.