Letting an AI agent design tables on a platform whose data structure changes with each business task

LayerX lets an AI agent design and create database tables from a conversation, but only after a person approves, and it may add tables and never delete them.

LayerX, a Japanese software company, published an account in Japanese of how its Ai Workforce product lets an AI agent design the tables it stores business data in. The author, suna, a software engineer there, explains which steps the agent takes and which a person still decides.

• Each customer of Ai Workforce, which runs AI agents for work such as invoice checks and contract review, needs different data fields, often discovered only after work begins.
• The old flexible storage made types hard to enforce, and the agent had to read every record to match invoices against purchase orders.
• In the new design, which works in a trial and has no release date, the agent proposes real tables, asks questions and creates them after a person approves.
• The agent may only add tables and columns, and each change to a record is shown first, approved by the user and reported back with its old values.

Teams giving AI agents write access to customer data can allow additions freely and hold back deletions until there is a way to restore the data. This is an inference and is not the author's claim.

5 November 2026: teams let AI agents gather the evidence and do the building, and decide in advance where a person must approve, check or judge.