How to use the Cubyts AI Sandbox?

Cubyts AI Sandbox

The Cubyts AI Sandbox is designed to give customers and prospects a safe, hands-on environment to explore and experience the platform’s most impactful features. It acts as a dedicated, time-bound agentic workspace where users can experiment with AI-driven capabilities, simulate real-world scenarios, and understand how Cubyts AI can accelerate their product development and delivery processes. By using the Sandbox, stakeholders can gain practical insights into the platform’s value before adopting it more broadly across their organization.

User Credentials

Access credentials for the agentic workspace in the Sandbox environment will be shared directly with the customer or prospect via email. These credentials provide the user with access to most of the platform’s features, enabling them to explore and experience the capabilities of Cubyts AI in a practical setting.

Integrations

The Sandbox agentic workspace is pre-configured with key integrations to help customers and prospects experience the breadth of Cubyts AI capabilities in a realistic setting:

1. Jira: Integrated with a sample project to enable discovery of process and feature drifts based on associated work items.

2. SOW: Connected to a sample Statement of Work (SOW) document to demonstrate how Cubyts AI identifies drifts from documented commitments.

3. Github: Linked to a sample code repository to showcase detection of code drifts against requirements and established technology standards.

4. Figma: Integrated with a project and design system to illustrate Cubyts AI’s ability to analyze requirement quality from design artifacts.

5. Google drive: Connected with sample documents to demonstrate analysis of functional/technical requirements and computation of code drifts.

Steps to use the sandbox:

Please follow the steps mentioned below to use the agentic workspace in the Sandbox environment:

1. Login: 

   1. Goto https://id.cubyts.com/login?o=true.

   2. Sign in using the user ID and password shared with you.

(Platform sign-in)

2. Flags Assistant:

   1. After login, you will be directed to the Flags Assistant.

   2. By default, it is filtered to showcase drifts in the active sprint executed by the team.

   3. (Note: You may deactivate this filter by disabling it, from the Flags filter view).
      

(Flags assistant)

3. Reports Assistant:

   1. Navigate to the Reports section from the navigation bar.

   2. Here, you can access all data assistants enabled for this workspace.

(Reports assistant)

4. Repository Assistant:

   1. Navigate to the Repository section from the navigation bar.

   2. Here, you can access all documents and associated contexts enabled for this workspace.

(Repository assistant)

Flags configured in the sandbox:

The Flags Assistant in the Sandbox environment is pre-configured with a set of key flags that demonstrate how Cubyts AI identifies and manages drifts across different dimensions of delivery.

Effort planning: Cubyts AI automatically discovers effort estimates for build work items based on established benchmarks (Note: Accurate effort predictions require that Cubyts AI has already built a reliable effort benchmark metric).

(Effort planner)

Milestone change discovery:  Tracks changes made in Jira during project execution and compares them against the original scope in the SOW. Flags deviations in terms of scope, time, and resources.

(SOW Milestone changes)

Requirements Quality Assessment: Extracts requirements from build work items and evaluates their quality against configured benchmarks, providing actionable recommendations for improvement.

(Requirements quality)

Drift between specifications and code: Analyzes the codebase to detect functional drifts between implemented code and the underlying specifications, and suggests corrective actions.

(Drift between specification and code)

Drift between technology standards and code: Reviews the codebase to identify deviations from established technical standards, providing recommendations to align code with best practices.

(Drift between technology standards and code)

Dependency checks in the codebase: Examines pull requests and branches to identify potential dependency mishaps that may arise due to recent changes.

(Dependency mishaps in the codebase)

Reports configured in the sandbox:

The Reports Assistant in the Sandbox is configured to showcase the end-to-end journey of code—along with impediments that may arise—spanning from delivery planning through to code execution. The following reports are configured and available:

Team reports: Highlight the quality of work produced by the team and the impediments encountered, enriched with data insights. These reports are based on streaming data extracted from Jira, GitHub, Google Drive, and Figma.

(Team reports)

Delivery reports: Assess the quality of delivery artefacts (e.g., Epics, Issues in Jira) and the associated impediments, drawing on streaming data from Jira, GitHub, Google Drive, and Figma.

(Delivery reports)

Code reports: Provide insights into the quality of code artefacts (PRs and Branches in GitHub), based on developer commits linked to work items extracted from Jira.

(Code reports)

Contract reports: Track and evaluate the status of Statements of Work (SOWs) imported into the agentic workspace, across multiple dimensions. 

(Contract reports)

Conclusion

The Sandbox serves as an interactive environment for customers and prospects to explore the full potential of Cubyts AI. By simulating real use cases, it provides a hands-on experience that demonstrates the value, flexibility, and impact of key features. This guided exploration not only builds confidence in the platform but also helps stakeholders envision how Cubyts AI can accelerate their own initiatives with clarity and efficiency.