NOTE: WATCH ALL THREE VIDEOS IN THIS MODULE
Data Models Details – Click Here
Welcome to the redSling Training Portal.
In this module, we explore how Orchestrator AI can help plan and coordinate the setup of an application from a single workspace, and how the requirements identified during the planning process can be translated into structured data models.
Using the existing SmartVendorWeb project, you will learn how to provide project requirements and supporting material to Orchestrator AI, analyze that information, identify unclear areas, review the AI-generated understanding of the application, and create a structured setup plan. You will also learn how to create a Vendor data model and configure the fields required to store vendor information within the application.
Orchestrator AI acts as a central planning and coordination layer for the project. It can read the material provided, understand project requirements, identify gaps, and coordinate the appropriate AI capabilities required for the next steps. Instead of focusing only on generating an individual screen or data model, Orchestrator helps look at the application as a whole and determine what needs to happen next.
Before starting the planning process, you can choose how Orchestrator should work with you. In Manual Assistive mode, Orchestrator acts when you explicitly ask it to perform an action. It can observe the project, provide suggestions, and recommend appropriate tools, while actions wait for approval. The platform also provides an Autonomous mode, where Orchestrator can carry out its plan as it progresses. However, actions such as deleting data, deploying the application, or changing backend configuration still require confirmation. In this example, the workflow is performed using Manual Assistive mode.
The next step is to provide Orchestrator with information about the application. Project material can include requirements documents, notes, screenshots, brand guidelines, spreadsheets, and other supporting files. Material can be uploaded directly, or the project can be described in your own words. A reference link can also be provided, such as an existing website, specification, or competitor reference.
For the SmartVendorWeb project, the supplied material describes a Smart Vendor and Procurement System. Once the project material has been added, the Analyze My Material action allows Orchestrator to read the information and begin building an understanding of the application.
Orchestrator identifies major business areas, workflows, data requirements, and other information required to plan the application. The material does not need to be perfectly structured. Rough notes can be used as a starting point while Orchestrator progressively builds a clearer understanding of the project.
After analyzing the material, Orchestrator identifies information that still needs clarification. The generated questions can cover areas such as the user roles supported by SmartVendorWeb, whether the Finance module is an internal application feature or whether vendor invoices need to integrate with an external accounting or ERP system, the application’s branding requirements, and the preferred authentication method.
These clarifying questions help reduce unnecessary assumptions about important parts of the application. Questions can be answered when the information is available, or skipped when they do not need to be resolved at that stage.
As the analysis progresses, Orchestrator builds a structured understanding of the project. The generated goals include centralizing the vendor and procurement lifecycle, allowing users to raise purchase requests, routing those requests for approval, generating purchase orders, registering purchased assets, managing incidents and maintenance, and processing vendor invoices.
Orchestrator also identifies the application’s major workflows and the data that needs to be stored. This provides an opportunity to review the AI’s understanding before continuing with the application setup.
Another important part of the planning process is reviewing the confidence levels provided by Orchestrator. Confidence can be shown across areas such as workflows, data, user roles, integrations, branding, and sign-in. A lower confidence level indicates that the AI had to infer information instead of finding it directly in the supplied material. This helps identify which parts of the project may require additional clarification before implementation.
The generated data requirements can then be reviewed. Based on the project material, the application requires entities such as vendors, purchase requests, approvals, purchase orders, products, incidents, maintenance records, and invoices. Orchestrator identifies these entities by connecting the business workflows described in the requirements with the information that the application needs to store. It also highlights areas where relationships between different entities will be required.
Once sufficient information has been gathered, Orchestrator prepares a Setup Plan for the application. The plan includes the application blueprint, new data models, required data tables, user types, and login flow. It also identifies an additional Approval Queue screen for handling procurement approval decisions.
Rather than recreating existing application screens, Orchestrator can recognize pages that are already available in the project and plan to enhance them where appropriate.
The Setup Plan also identifies proposed data tables. One of the proposed models is the product model, which includes information such as product name, SKU, description, category, preferred vendor, unit price, currency, and active status.
Another proposed model is the approval model. This model connects procurement requests with approvers and includes fields such as approval sequence, status, decision notes, submission date, and decision date.
Reviewing these proposed structures provides visibility into the data that Orchestrator intends to create before the setup actions are performed.
Once the proposed data structure has been reviewed, the required data tables can be created. Orchestrator executes the setup step and creates the required models for the project. Since the example uses an assistive workflow, the process remains visible and controlled, allowing the setup plan and individual actions to be reviewed. Completed actions can also be reviewed or undone when required.
Some setup steps may be skipped when they are not required or when sufficient information is not available. In this example, deciding the user types and generating the login flow are skipped. Orchestrator also identifies information that was not available from the supplied material. For example, because no brand document or logo was provided, branding requirements could not be fully defined.
The Activity section provides a history of the actions performed by Orchestrator, including material analysis, project understanding, clarifying questions, and setup actions. This provides visibility into how the application plan was created and what actions were performed during the process.
After understanding the application planning workflow, the module moves into the creation of a Vendor data model.
A data model defines the structure of the information that the application needs to store. For the SmartVendorWeb project, the Vendor data model is configured with the fields required to capture vendor information.
The first field is vendor_name, which uses the String field type and stores the name of the vendor.
Next, vendor_type is added as a String field to capture the type or category of the vendor. The status field is also configured as a String field and represents the vendor’s current status.
The Vendor model can also maintain a relationship with another data model. To establish this relationship, the created_by field is added and configured with the Reference field type. The user table is selected as the reference table, establishing a reference between the vendor record and the corresponding user.
The created_at field is then added with the Date-Time field type. This allows the model to capture the date and time associated with the vendor record.
The model also includes vendor contact information. The contact_person, contact_email, and contact_phone fields are added using the String field type. Together, these fields provide the basic contact information required for the vendor.
Once all the required fields have been configured, the complete Vendor model structure can be reviewed before creation. At this stage, the model contains vendor identification information, vendor type and status, a user reference, date-time information, and vendor contact details.
After reviewing the configuration, the Create Model action is used to create the Vendor data model so it can be used within the project.
By completing this module, you will understand how application requirements can be analyzed and organized through Orchestrator AI, how business workflows can be translated into data requirements, how confidence levels and clarifying questions can help refine the application plan, and how proposed data structures can be reviewed before implementation.
You will also understand the fundamentals of creating a data model in redSling, including configuring String, Reference, and Date-Time fields and establishing a relationship between the Vendor model and the User table.
Make sure you WATCH ALL THREE VIDEOS in this module to understand the complete workflow from application requirement analysis and planning through to data-model configuration and creation.
#redSling #redSlingTraining #OrchestratorAI #DataModels #VendorDataModel #SmartVendorWeb #NoCode #AIApplicationDevelopment #ApplicationDevelopment #DataModeling #AITools