GenAI Website

 

Access GenAI

Access the site here to use GenAI:

1. Provide a prompt, and

2. The system creates a microservice: a database, an API, and a multi-page application

   • Review the project - running screens, with data - and the database diagram
   • Iterate to get the requirements right

3. Download the project to continue development with rules and Python in your IDE

 

Web/GenAI is based on API Logic Server - docs here.

Click the image below to watch a 2 minute video:

About

1. Instant Working Software - Get the Requirements Right

Automation has turned your prompt into a microservice: a database, a working application, and a standard API.

It simply cannot be faster or simpler.

• Eliminate weeks to months of complex framework coding, db design, or screen painting.

• Far more effective than 'dead` wireframes, you can...

  • Collaborate with stakeholders using Working Software, live data

  • Iterate 15 times... before lunch.

2. Microservice Development - Declarative Rules and Python in your IDE

The speed and simplicity of AI, plus all the flexibility of a framework.

• Download the standard project, and customize in your IDE

• Use standard Python: e.g. provide Application integration (custom APIs and kafka messaging)

• Declarative security: configure keycloak authentication, declare role-based row authorization
  

• Declarative business logic: multi-table constraints and derivations using unique rules that are 40X more concise than code, extensible with Python
  

3. Deploy - Standard container, no fees, no lock-in

Created projects include scripts to automate docker creation, so you can deploy anywhere.

There are no runtime fees, no lock-in.

 

Develop

You can explore the created microservice on your own computer.

1. Download your project (customize in your IDE to add logic & security)

   • Observe the project is a set of models - not a huge pile of difficult-to-understand code

2. Executable Docker Image

   • The website provides a docker command to run the created project

 

Contact for full access

To create unlimited projects in your environment, contact ApiLogicServer@gmail.com for a free docker image, and project support.

The underlying services are also available in the genai CLI.

 

Prompt Design

Prompt design is "AI Programming". Consider the following.

 

Business Area

You can provide a very general prompt, for example:

• an auto dealership, or
• a restaurant

 

Database Oriented

Or, you can provide a specific prompt that identifies specific tables, columns and relationships (for example,

Database, API and Web App

Create a system for Customer, Orders, Items and Products

 

With Logic

Particularly interesting is that you can declare backend behavior with rules:

Database, API, Web App and Logic

Create a system with customers, orders, items and products.

Include a notes field for orders.

Use LogicBank to create declare_logic() to enforce the Check Credit requirement (do not generate check constraints):
1. Customer.balance <= credit_limit
2. Customer.balance = Sum(Order.amount_total where date_shipped is null)
3. Order.amount_total = Sum(Item.amount)
4. Item.amount = quantity * unit_price
5. Store the Item.unit_price as a copy from Product.unit_price

You can verify this by altering a sample order/item with a very high quantity, and verifying the credit limit is checked. (Note this is not trivial - 3 table transaction.)

Note: at the time of this writing, the sample data sometimes does not totally reflect the derivation rules. This is under investigation. You can sometimes resolve this using iterations, eg, please initialize the customer balance properly.

 

Iterations

You can iterate your prompt to include more tables etc, while preserving the design you have already created.

• This enables you to break your system down into a set of "Use Cases", solving one at a time, and integrating back to the others.

 

Limitations

The created systems are basic database applications, not completed systems with sophisticated functionality such as images, custom screens, etc.

You can "build out" the project by downloading it and using your IDE with Python and rules, or perform the same functions using Codespaces (a browser-based version of VSCode - a link is provided for this).

 

Context

This explains the premise behind GenAI, and how it fits into a project life cycle.

 

Why GenAI

A common project experience is:

• Weeks (or months) to get running screens
• Then, when Business Users explore the screens, it becomes clear there were basic misunderstandings

Which leads to our premise:

1. Instant screens (Agile "Working Software"); collaborate, and iterate to get the requirements right

2. Kickstart the project with a solid backend - a Database, API and Logic from declarative / Natural Language models

   • Logic in particular is key: security and multi-table derivations / constraints constitute as much as half your project: declarative rules make them 40X more concise  

   • The remaining logic is built in your IDE using standard Python, providing the speed and simplicity of AI - including logic - with the flexibility of a framework

 

Project Life Cycle

As shown in the Life Cycle outline below, this approach is complementary to your existing UI Dev tools such as UI frameworks or Low Code Screen Painters.

• In addition to existing UI Dev tools, please explore automation support for Ontimize.

As illustrated by the green graphics below, we seek to provide value in the Inception Phase (get the requirements right), and for backend development.