Legal Case Management Platform

Our primary goal was to transition the codebase from ASP.NET Forms to Angular 11 without causing any disruption to the client’s critical system, which serves the government and legal sectors. To achieve this, our experts rewrote the legacy codebase, module by module, ensuring seamless integration with Angular while enhancing performance using TypeScript.

A crucial aspect of the migration involved integrating a wide range of UI components from the legacy system into the new framework, such as forms, buttons, modals, settings pages, maintenance pages, search functionality, user profile pages, services pages, menus, and navigation. By carefully incorporating these UI elements, our development team ensured that the system retained its core features while leveraging the capabilities of the new framework.

While migrating the legacy code to the new framework, our team encountered legacy API endpoints on the backend that required careful handling. API endpoints serve as the entry points for communication between the frontend and backend, enabling data retrieval through requests.

We meticulously evaluated each endpoint to determine the most appropriate approach for handling it within the new Angular framework. For some endpoints, we replaced outdated implementations with new ones, while for others, we replicated them and made necessary adjustments to align with Angular’s framework. This approach enabled us to use the existing backend infrastructure while ensuring compatibility with the new front-end framework.

The existing search feature, built with ASP.NET Web Forms, proved slow and inefficient, hindering effective data retrieval for prosecutors and defenders. To tackle this challenge, our team developed a new search engine from scratch, using modern technologies and a new design. We focused on simplifying the search process for users, allowing them to input queries, apply filters, and easily navigate results.

We integrated the Azure Cognitive Search service into the system, establishing a dedicated database known as the Index within the Azure environment. Unlike before, where the entire database was queried, the new search system relies on this Azure Index database. It contains a carefully filtered subset of records tailored to match user preferences and search patterns. This ensures that only relevant records are retrieved, optimizing search efficiency.

The revamped search functionality has delivered remarkable improvements in system performance. Previously lengthy search times have been reduced to an impressive 1–2 seconds, giving users swift access to the information they need.

Our customer faced a challenge handling lengthy video recordings, such as those from police body cameras, which needed modifications like trimming, highlighting specific elements, and blurring faces or objects. Previously, such modifications were made using separate third-party applications, potentially violating legislation in case of data leakages. To address this challenge, Leobit experts suggested consolidating audio/video editing tasks within the client’s ecosystem to ensure compliance with legislation on sensitive information.

Our solution was to develop a comprehensive online video editor. We used modern technologies like .NET 8, Docker & Docker compose, React, and Azure Blob Storage. By integrating with Azure AI Video Indexer, we enabled users to extract valuable insights from videos, including automatic translation and multi-language support.

The media editor’s key features include face and object blurring, annotation drawing tools, and AI-powered caption extraction and editing, allowing users to perform all necessary actions seamlessly within the platform.

Following the latest application upgrade, our team introduced a fully automated flow. Now, once a video is uploaded, it is automatically processed by Azure AI Video Indexer, which generates a transcript without user involvement. The system then applies additional formatting and security rules, including masking sensitive data using regex, converts the output into a structured document, and stores it in the main application. At the same time, all transcripts are indexed using AI-powered search, allowing users to quickly find relevant content across videos.

The impact of this upgrade became clear immediately after release. Just after the feature launch in April 2025, the media editor handled around 200–300 videos. After the upgrade, usage rapidly scaled to 10,000–15,000 video uploads per day, with each video averaging about one hour in length. To support this growth, the system relies on Azure auto-scaling, handling approximately 15 terabytes of data daily and reaching around 5.5 petabytes annually. This shift turned the media editor into a high-load, production-critical system capable of processing massive volumes of video efficiently and securely.

We developed a cross-platform mobile and desktop application specifically designed to provide victims with access to case information and facilitate communication with their legal counsel. This application functions as a separate microservice, communicating with the main application through an API. Our main challenge lay in seamlessly integrating this microservice with the core application while ensuring the secure storage of data in the primary database.

To address this challenge, we employed a multitenancy approach, enabling the application to connect to various client environments in production. This architecture ensures flexibility and scalability, accommodating different client needs effectively.

The app gives victims access to the essential case details, including evidence, and facilitates seamless communication through feedback mechanisms. Previously, feedback collection relied on disparate methods like Google Forms, emails, or letters. Now, victims receive timely notifications prompting them to provide feedback directly within the app.

Additionally, the app features an embedded calendar displaying court dates, locations, and attendance requirements, offering victims clear visibility into their legal proceedings. Furthermore, users can access documentation FAQs and even send subpoenas to court, all within the app interface.

On the administrative side of the main app, we implemented functionalities allowing administrators to manage user texts, FAQs, victim profiles, document generation, and access permissions. Moreover, administrators can configure session durations, ensuring security by setting expiration limits for access.

Since its release in April 2024, the portal has seen steady adoption, reaching approximately 9,500 users by April 2026. Such growth confirms its value as a reliable communication channel between victims and legal teams.

The Court Face Sheet is the main working page for prosecutors during court hearings. It includes over 30 modules and aggregates information from multiple parts of the system.

Previously, this information was scattered across different pages and tabs, requiring users to search and navigate under time pressure. The client had an outdated version of the Court Face Sheet built with Windows Web Forms. We developed a new page that brings everything together into a single interface using modular blocks.

Because of its critical role, the feature could not be released incrementally. It had to be delivered as a complete solution from day one, while fully meeting accessibility requirements. From a technical perspective, the court summary page was implemented as a standalone Angular module with a dedicated back-end module.

One of the main challenges was integrating data from legacy pages. The team used a hybrid approach:

• Some logic was rewritten from scratch
• Some parts were encapsulated from existing functionality
• Additional features were extended or newly developed

To support gradual adoption, the team implemented feature toggles. This allowed users to switch between the old and new versions while both existed independently. Over time, the legacy version is being phased out.

Police officers needed a reliable way to track court appearances, confirm attendance, and manage scheduling conflicts, but the existing system didn’t support how they actually worked. It was designed for desktop use only, not responsive, and difficult to use on mobile devices. For officers who spend much of their time in the field, this made the tool inefficient and, at times, unusable.

Instead of trying to patch the old system, the Leobit team chose to rebuild it with mobility and simplicity as core principles. On the front end, the team reused a custom component library that we built for the court summary page to ensure consistency and speed up development. On the back end, we adapted an existing .NET system to accommodate new data structures, while features such as search and filtering were developed from scratch. As a result, the module fits naturally into daily workflows and works seamlessly across devices.

The Leobit team developed and implemented a custom AI agent to help users quickly navigate large volumes of legal data and extract key insights without manual effort. Instead of reviewing multiple documents and screens, users can generate concise summaries of cases on demand. To make this possible, the team built a custom AI-driven solution powered by Azure OpenAI, integrated into the existing system through a .NET and ASP.NET Core Web API back end. To ensure accurate and relevant summaries, case data stored in Azure Cosmos DB is processed and structured before being sent to the AI model.

On the front end, the team designed and built a dedicated Angular interface that allows users to easily request and view summaries within their workflow. The solution was integrated with existing media and document handling pipelines, including Azure Blob Storage, so both textual and transcribed content could be used as input. By combining modern cloud services with a clean architectural approach based on Onion architecture, the team delivered a feature that significantly improved efficiency. Tasks that previously required several hours of searching through case materials can now be completed in less than one minute, allowing users to focus on decision-making rather than data gathering.

As the platform evolved to include multiple modules covering various aspects of case handling, regression testing became a significant operational bottleneck. The process became increasingly complex and time-consuming, requiring over 40 hours per testing cycle when performed manually.

To address this challenge, Leobit’s QA team proposed implementing automated testing for the application’s core functionality and critical end-to-end user flows. Using Mabl, a cloud-based intelligent test automation platform, the team automated key scenarios across high-impact modules to ensure consistent quality and faster release cycles.

The QA team worked closely with the development team to ensure that all test cases were aligned with the functional requirements. They prepared and optimized 281 automated test scripts, each taking an average of 8 hours to set up and fine-tune.

The automated tests were deployed on a dedicated remote server and integrated into a nightly run schedule, which allowed us to provide continuous feedback and early detection of issues. This approach helped us significantly reduce testing time and improve reliability and coverage with every release. This scalable testing framework can grow alongside product expansion.