inbybob_

Warehouse Management APP

• Team: 1 functional architect, 2 System analysts, 2 frontend developers, 2 backend developers, 1 designer, 2 QA engineers
• Technologies: Java, Spring Boot, Hibernate, Redis, React, Postgres SQL
• Duration: 12 months

Problem

The business faced problems with storing and accounting for a large volume of goods in warehouses:

• A large volume of goods in each warehouse;
• Difficult to track the constant movement of goods and changes in their quantity;
• Need to consider standards for storing different types of goods.

Solution
To solve this problem, it was decided to create an application that would help in warehouse accounting. For this purpose, based on mobile scanners like Zebra, it was decided to implement a mobile application for warehouse accounting.

Resolved Issues: solved the problem of accounting for warehouse good,  reduced the time spent on inventory in the warehouse, reduced the time to find the necessary goods in the warehouse from several hours to several minutes, reduced the time to issue goods to the customer from tens of minutes to 5, created a unified database for storing warehouse goods

Voice Interface for Warehouse Employers

• Team: Project manager, Solution architect, Analyst, 2 Backend developers, and 2 QA Engineers
• Technologies: Java 11, Spring Boot, PostgreSQL, JSAPI
• Duration: 6 months

Problem

This project involved developing a voice interface for data collection terminals used by a warehouse with a significant number of foreign workers. The aim was to address errors caused by language barriers and difficulty reading small screens.

Solution
Introduction of the voice interface led to an average 7% increase in worker efficiency.

Error rates dropped by 10%.

Facet Search for E-Commerce

• Team: Product/Project Manager, Business Analyst, 4 Backend developers, 3 QA Engineers, DevOps Engineer
• Technologies: ELK, Java 11, Kotlin
• Duration: 12 months

Problem

At the outset, the client highlighted the issue of the uncontrollability of the SaaS solution provided by the contractor. The availability of all sales channels directly depended on the search functionality, as products on the website, in buyers’ mobile apps, in offline store inventories, and in sellers’ mobile apps also relied on this solution.

At the start, the search index update frequency was once every few days, which significantly delayed the appearance of new products on the website and other channels.

Off-the-shelf solutions led to the problem of vendor lock-in, unlike solutions built on open-source technologies.

Solution

POC
During the Proof of Concept (POC) phase, our team was tasked with developing a solution for the website that would be at least as good as the current one. This was to maintain the existing user experience and assess how much it could be improved compared to the vendor’s SaaS solution.

Over four months, a mixed team consisting of a business analyst from the client side, a project manager, a data analyst, three developers, one QA engineer, and a DevOps engineer developed a solution based on Java 11, Kotlin, and the ELK stack (Elastic Search for search, Logstash for logging, and Kibana for analytics and monitoring).

Key metrics at the end of the POC phase were Precision and Recall to ensure that the new indexing algorithms did not negatively impact performance.

The POC results demonstrated that the metrics were maintained at the same level as before, and we achieved daily updates to the primary search index.

2 Phase
During the second phase, the development team was expanded to four developers, and the testing team grew to three QA engineers.

A separate index was created for autocomplete suggestions, along with several reference guides to manage ranking for specific categories of promoted and recommended products. Additionally, manual boosting was implemented for new, popular items to address cases where a product had not yet accumulated a “popularity or sales weight.”

Improved analysis of the search query for the presence of phrases typed with an incorrect layout. Extracting product attributes from a search query (conceptual search).  Improved indexer based on In-Memory solution with support for partial indexing. Collection of statistics on search results based on ELK.

The team achieved an increase in the average check for sessions from a search by 1.5%, and the share of sessions using the search by 7 percentage points.

QA Audit for EMEA bank

Stages of Auditing a Product Quality System

SDLC processes built without a dedicated testing team are inside a large banking project that has been running for five years and consists of numerous microservices and integrations with other systems.

Success Metrics

• COBIT/ITIL/RUP/ISO/CMMI/PMBOK maturity level parameters and points to improve to reach target level
• Measurement of development team performance as a current state with points to improve: some text
• ROI for Autotests - time for development autotest / time for manual testing
• Time to fix defects
• Time to market (feature to production)
• Average time for regression testing

Recommendations for establishing a QA process in a project without dedicated testers and using .NET and Oracle technologies:

1. Establishing a Testing Structure
Recommendations:

• Appoint responsible individuals in the development team to coordinate the testing process.
• Develop a basic test plan that includes key components:some text
• Functional testing.
• Integration testing.
• Performance testing (if critical).
• Both manual and automated testing.
• Implement testing practices at every stage of development (starting with unit tests).

2. Test Automation
Recommendations:

• Unit testing: Set up unit testing to test individual modules of the system.
• Integration testing: This is especially important for microservices and interactions with databases (Oracle).
• Contract testing: For microservices, ensure they interact correctly via APIs.

Tools for .NET:

• NUnit or xUnit: writing unit tests in .NET.
• Moq: To create mocks and stubs for testing individual modules.
• SpecFlow: A tool for writing tests using BDD (Behavior Driven Development), useful for integration testing.
• Selenium: For automating user interface testing (if the project includes a web application).
• RestSharp or Postman: For testing microservice APIs. Postman can also be used for API test automation.

3. Tools for Database Testing
Recommendations:

• Test the performance of queries to Oracle, especially if the system needs to handle large volumes of data.
• Check data integrity during integration testing.

Tools for Oracle:

• SQL Developer: A tool for testing SQL queries, analyzing query performance, and testing stored procedures.
• TSQLt: A framework for testing databases that can be adapted for testing SQL queries and procedures in Oracle.
• Oracle SQL Performance Analyzer: To test query performance and assess how code changes impact database performance.

4. CI/CD and Testing at All Stages
Recommendations:

• Set up a CI/CD process with automatic test runs at every stage of development. For example, run unit tests with every commit, and run integration tests at the build stage.
• Use code reviews as part of the QA process. Integrate code checks for compliance with standards through static code analyzers.

Tools:

• Jenkins or Azure DevOps: Support .NET and Oracle and can help automate test runs.
• SonarQube: A tool for static code analysis, which can integrate with CI/CD to control code quality and test coverage.

5. Test Case Documentation and Management
Recommendations:

• Establish a system for test case management and documentation of test scenarios. This will help track which features have been tested and which need additional attention.
• Introduce a practice of regression testing to check key functionalities after each change.

Tools:

• TestRail or Zephyr: Tools for managing test cases and test documentation. They can also integrate with Jira or other project management systems.
• Azure Test Plans (in Azure DevOps): Supports creating test cases, maintaining test documentation, and integration with CI/CD processes.

6. Manual Testing and Basic QA Processes
Recommendations:

• Implement a practice of manual verification of new functionality until it is automated.
• Defect tracking: Ensure that every identified defect is documented in a bug tracking system (e.g., Jira) and that tests are created based on it.

Tools:

• Jira or Azure DevOps: For managing tasks and bugs.
• Exploratory Testing: Conduct exploratory testing to identify hidden defects and bugs.

7. Monitoring and Production Support
Recommendations:

• Set up a system for monitoring performance and identifying issues in production (e.g., microservice failures, and database crashes).
• Implement a practice of regular stress testing to ensure the system can handle loads.

Tools:

• Grafana + Prometheus: For monitoring the health of microservices and the database.
• JMeter or k6: For load testing to ensure the system can handle large volumes of requests.

Summary of Implementation Plan:

1. Establish a basic testing structure
2. Set up unit testing and microservice testing
3. Integrate tests into CI/CD
4. Document and manage test cases
5. Monitoring and control in production

Migrating custom functionality from SAP Commerce Product Cockpit to Backoffice PCMT

• Team: 3 Backend Developers, 1 Team Lead, Quality Assurance Engineer, Business Analyst, Project Manager
• Technologies: SAP Commerce Cloud (Hybris), Angular, Solr, Spring, Java, Microsoft SQL Server
• Duration: 6 months

Challenge

SAP Commerce Cloud (Hybris) released a new version, deprecating support for the client’s existing version. This posed a challenge for the client, as the Product Cockpit (PC) used for catalog and product management was no longer supported. The new Backoffice PCMT (Product Content Management Tool) introduced structural changes, requiring a full migration of custom functionalities.

The key challenges includes:

• Feature Compatibility – Many customized features in Product Cockpit were not available in Backoffice PCMT.
• Process Optimization – The client’s workflows were built around the legacy system, requiring process adjustments.
• Business Continuity – The migration had to be seamless with minimal downtime and zero disruption to ongoing operations.

Feature gap analysis & Process optimization
Objective: Identify gaps between the old and new systems while streamlining processes.

• Conducted a full audit of the client’s customized Product Cockpit features and compared them against native capabilities of Backoffice PCMT.
• Identified deprecated features, missing functionalities, and new opportunities within the updated system.
• Collaborated with business stakeholders to reassess existing workflows, eliminating redundant steps (e.g., simplifying promotions setup, catalog, and product management).
• Provided recommendations on adapting to Backoffice PCMT and outlined necessary customizations to maintain operational efficiency.

Development & Migration strategy
Objective: Migrate essential features while ensuring full compatibility and scalability.

• Developed a prioritized migration roadmap, focusing on:

1. Critical functionalities (e.g., product enrichment, attribute management, promotions, versioning).
2. High-impact automation for catalog workflows.
3. User access and permission management to align with Backoffice PCMT structure.

• Re-implemented key customizations, ensuring they were compatible with the new architecture.
• Built custom extensions where the new system lacked native features to replicate legacy behaviors.
• Refactored existing integrations to align with Backoffice PCMT’s data model and API structure.

Testing & Validation
Objective: Ensure a smooth transition with zero operational downtime.

• Developed comprehensive test cases for functional, integration, and user acceptance testing (UAT).
• Conducted parallel runs where both Product Cockpit and Backoffice PCMT operated simultaneously to validate accuracy.
• Engaged end-users early in the testing phase to gather feedback and refine UI/UX adjustments.
• Ensured data consistency and backward compatibility during the migration process.

Deployment & Rollout
Objective: Deliver a seamless go-live experience with minimized risk.

• Implemented a rollback strategy to mitigate potential risks.

Provided onboarding sessions and user training to ensure smooth adoption of Backoffice PCMT.

Results

• 100% successful migration of all essential functionalities to Backoffice PCMT.
• 40% reduction in manual catalog management tasks due to workflow optimization and additional integrations.
• Enhanced performance & UI experience, leading to faster product setup and approvals.
• Zero downtime & seamless transition, ensuring business continuity without disruption.