Use Case Development

With the MVP design finalized and all pre-development checkpoints completed, the use case enters the Development phase in Calibo’s Sandbox. This is where actual implementation begins—application code is developed, data pipelines are built, and analytics models are engineered to create a robust, data-driven solution.

Application Development

Developers begin by selecting the required tools and technologies—such as Angular, React, Node.js, Cucumber, and SonarQube—from the pre-approved tech stack defined in the policy template applied to the use case. Based on these selections, dedicated source code repositories are automatically created in the configured version control system (such as GitHub, GitLab, or Bitbucket). This automation offloads repetitive setup tasks, allowing developers to focus entirely on writing and testing code.

All development activities are fully traceable, with Jira used to track epics and user stories and Confluence providing access to technical documentation and design artifacts. This integrated setup ensures alignment between development, design, and business goals.
Data Engineering and Pipeline Development

In parallel, data engineers configure pipelines to ingest data from disparate sources using tools like Databricks, or Snowflake. They execute ETL (Extract, Transform, Load) jobs, apply data quality rules, and ensure that the resulting datasets are reliable, clean, and analytics-ready.

Data analysts further enhance these datasets by performing operations such as join, union, aggregation, filtering, and enrichment. They apply domain-specific rules and business logic to prepare structured data models, which are then handed over to data scientists.
Machine Learning and Analytics

Data scientists use integrated tools like JupyterLab to apply machine learning algorithms (like Random Forest Classifier) and build custom models for specific use case requirements. They run analytics pipelines to perform feature engineering, model training, evaluation, and versioning. Once validated, the models are made available to Business Analysts or BI developers to consume and visualize insights through dashboards or embedded APIs.

Thanks to the structured setup from earlier phases—including a refined backlog of user stories, approved design artifacts, and pre-configured DevOps toolchains—development proceeds with clarity and minimal friction. Product Owners, Developers, QA leads, and Engineering Managers collaborate through agile sprints to ensure steady progress and incremental value delivery.

While developers focus on building the current use case, Product Owners can begin preparing the next prioritized use case—enabling continuous delivery across the portfolio.

Goals	Outcome
Enable seamless setup of development environments through predefined policy templates, platform configurations, and data pipeline workflow. Automate the creation of dedicated source code repositories (GitHub, GitLab, etc.). Ensure traceability and collaboration through Jira and Confluence Facilitate developers to begin coding and test code directly in the self-service platform. Provide direct access and management of source code within the self-service platform Support iterative analysis and agile refinement, empowering developers to raise design discrepancies during sprints for validation and updates.	All foundational artifacts—architecture, UI/UX, and data models—are finalized, accessible and ready for development. Development frameworks, tools, and platforms (e.g., Angular, Node.js, GitHub, SonarQube). Development repository created like source code/stories, environments configured, and policy template applied to the product being developed. Connections established with project tracking tools like JIRA and documentation tools like Confluence. Developed and tested use case on Sandbox environments (Dev to QA)

Goals

Outcome

Enable seamless setup of development environments through predefined policy templates, platform configurations, and data pipeline workflow.
Automate the creation of dedicated source code repositories (GitHub, GitLab, etc.).
Ensure traceability and collaboration through Jira and Confluence
Facilitate developers to begin coding and test code directly in the self-service platform.
Provide direct access and management of source code within the self-service platform
Support iterative analysis and agile refinement, empowering developers to raise design discrepancies during sprints for validation and updates.

All foundational artifacts—architecture, UI/UX, and data models—are finalized, accessible and ready for development.
Development frameworks, tools, and platforms (e.g., Angular, Node.js, GitHub, SonarQube).
Development repository created like source code/stories, environments configured, and policy template applied to the product being developed.
Connections established with project tracking tools like JIRA and documentation tools like Confluence.
Developed and tested use case on Sandbox environments (Dev to QA)

Once development activities are complete and the solution passes QA and quality gates, the use case is ready to transition into the Deployment phase.

Checklist for Readiness

The Use Case Development Checklist plays a critical role in translating approved designs into robust, scalable, and deployment-ready solutions. It provides a structured validation framework to ensure all technical, operational, and collaborative elements are properly set up before the build progresses to the deployment stage.

Sl. No.	Item	Status (Not Started/In Progress/ Completed)
1	Tech stack selected based on the solution design	Not Started
2	GitHub/GitLab repositories auto-created for frontend and backend technologies and data pipelines	Not Started
3	CI/CD tools configured (for example, Jenkins, GitHub Actions)	Not Started
4	JIRA board set up for user stories and connection to JIRA established via Calibo Sandbox	Not Started
5	Connection to Confluence established via Calibo Sandbox	Not Started
6	Finalized UI/UX and data models uploaded	Not Started
7	Use case developed and tested in the Sandbox environment (Dev → QA)	Not Started
8	Code quality checks (for example, SonarQube) run through automated CI/CD pipeline	Not Started
9	Feedback captured and resolved through sprint cycles	Not Started
10	Version control implemented	Not Started
11	The use case marked 'Ready for Deployment' in Calibo Sandbox	Not Started

PRO TIP:

Integrate your JIRA, Git, and Confluence early. Linking user stories, code commits, and documentation in real time boosts traceability and simplifies future audits or handoffs.

Advance Bank Bringing Sentiment Analysis Engine to Life

With the MVP designs approved, architecture finalized, and all pre-development checkpoints complete, Advance Bank was ready to move into the Development phase, transforming vision into working code, backed by version control, automated CI/CD pipelines, and seamless tool integration in Calibo Sandbox.

This wasn’t just about writing code—it was about creating an execution-ready, collaborative environment where cross-functional teams could build, test, and iterate without bottlenecks.

Use Case: Sentiment Analysis of Customer Product Reviews

Goal: Automatically classify customer feedback—positive, negative, or neutral—and visualize these sentiment trends through real-time dashboards to improve product experience and responsiveness.

Step	Personas Involved	Description
Environment Setup	Product Owner Developers DevOps Engineer	Selected the approved tech stack (React, Python/FastAPI, Azure NLP, PostgreSQL) based on the policy template.
Repo Creation	DevOps Engineer Developer Leads Platform Engineer	GitHub repositories were auto-created for frontend, backend, and pipeline components.
Frontend Development	Front-End Developers UI/UX Designers	Built an interactive dashboard to display sentiment trends. Implemented filters for product, time range, and sentiment category. Integrated design changes from Figma and adhered to component-level guidelines defined in the MVP.
Backend Development	Back-End Developers Solution Architect	Developed RESTful APIs to expose sentiment scores. Created endpoints for dashboard queries, feedback submission, and system health. Implemented pagination, sorting, and basic auth middleware.
Data Pipeline Setup	Data Analysts Data Engineers	Performed joins with product metadata. Applied aggregation, filtering, and domain-specific rules. Prepared structured, clean data for use in dashboards and modeling.
Modeling & Analytics	Data Scientists	Used JupyterLab to apply predefined ML algorithms like Random Forest. Conducted feature engineering, model evaluation, and versioning. Updated model weights based on feedback loops. Delivered outputs for consumption in dashboards.
Sprint Demo & Feedback Loop	Product Owner QA Lead Front-End & Back-End Developers Data Scientist Business Analyst	Mid-sprint, a demo revealed a gap—the dashboard was charting data but not aligning with the KPIs defined during design. The product manager quickly updated the Figma file, realigning the visual expectations. At the same time, the QA team flagged that the sentiment model wasn’t handling sarcasm in reviews effectively. The backend lead requested updated model weights and scheduled an update. Meanwhile, the business analyst made sure that all changes—design tweaks, new model inputs, and test logs—were updated in Confluence and reflected in JIRA tickets.
Testing & QA	QA Lead Developers Product Owner	Functional, integration, and data quality testing performed. SonarQube code quality thresholds enforced. All stories marked as Done in Jira.
End-of-Sprint Readiness	Product Owner DevOps Engineer Product Release Manager	All code and artifacts reviewed. CI/CD pipelines validated. Artifacts versioned and tagged. The use case marked “Ready for Deployment.”

By the end of the sprint, the use case was no longer a plan. It was a product taking shape—tested, traceable, and team-aligned. The use case is now ready to transition to the Deploy phase.

What's next? Use Case Deployment