API Maker
The backend engine for the AI era
Generated APIsEpisode 6 4:19 12 chapters October 2026
Every table answers REST APIs without a line of code: seventeen operations on MongoDB and fourteen on SQL, in two families, SCHEMA and GEN, with one envelope for success and error. Take the tour from the panel, from a terminal with curl, and in the Swagger docs of an application.
In the playlists The complete course Generated APIs
Every table answers APIs, without a line of code : seventeen operations on MongoDB, fourteen on SQL, in two families, with one envelope. In this video, a tour of them, from the panel, a terminal and Swagger.
An API for a table usually means a controller, a service and a repository, table after table, an answer shaped differently by every team, and API docs written by hand, late and out of date. With API Maker, every operation of every table is there, every API answers the same envelope, and the Swagger docs are written from what an application may call. In this video : the APIs of a table, SCHEMA and GEN, one envelope, an error, a call from a terminal, and Swagger docs for one application.
Instance API, mongodb, catalog, products : thirty four APIs. Seventeen operations, get all, get by id, save, update, master save, count, distinct, aggregate and more, under SCHEMA, and the same seventeen under GEN. On SQL, crm, customers : twenty eight, fourteen of each. Replace by id, array operations and aggregate are for MongoDB.
A SCHEMA API reads the schema of the table : values converted, rules checked, relations followed. A GEN API leaves the schema out : what you send is what the database gets. Same operations, same addresses, under slash api slash schema, or slash api slash gen.
API Testing, products, Get all, SCHEMA. find : a price of thirty five, written as a text. select : the name and the price. Two products : the schema made the text a number. Around them, the envelope : success, statusCode, and data.
The same call on GEN Get all, the same find. Nothing : the text stays a text, and no price is a text. The same envelope, its data empty.
Another engine, PostgreSQL, the customers, and a mistake : an id which is not a number. Four hundred : success false, and errors, one entry per problem, with its type, its field and its message.
From any client, the same call : an address, and the token of an API user in x-am-authorization. The same envelope, from a terminal. Without the token : four hundred and one, in the same envelope again.
API User Permissions, Support Desk, the application of video twenty four, Edit. Enable Swagger Docs, and a random token : the end of an address only you know. Update. Swagger UI, that address, Explore : exactly the three APIs its group opens. Get by id, with its id, ready to try.
Custom APIs and system APIs answer the same envelope. The status is in the body too : two hundred, two hundred and one for a save, four hundred, four oh one, four oh three, five hundred. Groups decide which application calls which API : video twenty four. And the Swagger docs follow the groups : change a group, the docs change.
One envelope, for every API of every table. Errors field by field, the same on every engine. So a client, written by you or by an AI assistant, handles every API the same way.
To sum up. Every operation of every table, without code. SCHEMA reads the schema, GEN leaves it out. One envelope, success or error. Swagger docs for each application. Next : the read APIs, one query language for six databases. API Maker. Build faster, spend smarter, deliver better.
Length 5:01Read APIs
Length 3:25Write APIs
Length 3:16Master save
Length 8:43Deep populate