TCS' AI-Assisted Application Modernization Framework - TCS AI Catapult Forward : 2-Hrs Briefing and Demonstration

"TCS AI Catapult Forward" provides the ability of choosing the right reference architecture of application and then enables the application designer to decompose the projects into multiple fragments of build work. Each build item then uses the framework’s built-in automation and user provided prompts, and context-aware references to create code for that object. The user can then finalize the code after reviewing. The framework also automates unit testing, devops pipeline creation and IaC development using Azure OpenAI. This approach fosters high-speed application modernization, boosting productivity and efficiency for developers.

The framework utilizes the design outputs to create a project template with prepopulated content generated by Azure OpenAI’ s generative code snippets. The input for the framework includes.

Project Technology: The technology stack or platform on which the application is built. Architecture type: The architectural style or pattern used in the application. API Definition: Specifications for the application's APIs, including the expected inputs and outputs. Business Functionality: Descriptions or explanations of the application's business logic, written in natural language. Data Model / Objects: The data model or object structure required for the application in natural language. Azure resources: The Azure resources needed to deploy and run the application.

Based on these inputs, the framework automates the generation of Azure developer template project that includes the following components:

1. Infrastructure: The framework leverages its own template library and Azure OpenAI to generate Infrastructure as Code (IaC) using tools like Terraform or Bicep. It configures the necessary Azure resources based on the provided information.
2. Application: TCS's inhouse framework is used to generate standardized project scaffolding. Additionally, Azure OpenAI’ s generative capabilities are employed to transform the business logic into functional code using prompt engineering.
3. Database: The framework leverages Azure OpenAI to generate context-aware table schemas and static data for the database.
4. Testing: Azure OpenAI, in conjunction with TCS's inhouse framework, helps generate the test project and unit test cases by understanding the context and requirements.
5. CI/CD: Finally, the framework assists in building CI/CD pipelines YAML files using Azure OpenAI’ s capabilities.

In summary, the framework uses Azure Open AI and TCS's inhouse framework to automate the generation of a project template, infrastructure code, application code, database schema, testing components, and CI/CD pipelines, resulting in accelerated application development and deployment.

Once the project template is generated, it can be utilized for further development activities, including refining, and optimizing the generated code or writing new code. To speed up the modernization process, an AI Paired programming approach can be adopted by integrating GitHub Copilot into the development workflow.