Operating ProcedureExecution layer

Fusekit Client App Deployment (Delivery)

A sequence with 6 steps: Create a new app in the Fusekit dashboard using the AI builder.

By InnovaAI ResearchPublished

What are the steps?

sequence

Fusekit Client App Deployment (Delivery)

  1. 01

    Create a new app in the Fusekit dashboard using the AI builder

    Describe the client's internal workflow in natural language. The AI builder generates a working draft that you can refine before connecting data sources.

  2. 02

    Connect the client's approved data source via the Data Connectors panel

    Select from PostgreSQL, MySQL, MongoDB, S3, Airtable, SQL Server, Snowflake, or Google Sheets. Note that operations are read-only, so plan the app's functionality around viewing and reporting rather than writing back to the source.

  3. 03

    Configure role-based access in the Identity settings

    Set up team member permissions so each client user sees only the data and actions relevant to their role. Fusekit handles authentication automatically, so you don't need to build login screens.

  4. 04

    Deploy the app to a protected URL using the Publish button

    Fusekit assigns a protected URL with built-in work sign-in. You can also use your own web address if the client wants a branded domain.

  5. 05

    Validate the app with the client and iterate on feedback

    Walk through the deployed app with the client, collect feedback, and return to the AI builder to make adjustments. The runtime updates in realtime, so changes appear immediately.

  6. 06

    Document the app logic and deliver a recorded handoff walkthrough

    Record a screen capture of the app's key features and data flows. Provide this along with a written summary of the app's logic and any limitations (e.g., read-only data).