Important Note: In order to comply with many of our clients confidentiality agreements, in some cases it is not possible for us to mention their names or trademarks for commercial purposes. For this reason, you may find wording such as “Client Name Confidential” or generic references to their industry rather than the client’s specific name. This choice reflects our commitment to protect the confidentiality and mutual trust that binds us to our partners.
Overview
The client is a leading banking and finance organization. Its finance and investment division who was looking for a digital platform that could visualize data and insights from multiple sources, including third parties.
The project involves the implementation of a complete data capture, processing and visualization flow, in which the processing phase applies complex logic to highly sensitive information. This phase is enhanced by artificial intelligence, which enables the generation of predictive insights and future projections.
Data inputs include credit scores, capital management metrics, market dynamics and financial behaviors, among others, providing in-depth and nuanced insights.
Once processed, data are presented via a customized, intuitive dashboard based on role-based access controls, fully compliant with industry regulations. The system is integrated with SPID (Public Digital Identity System) and supports multifactor authentication mechanisms.
After a careful evaluation phase and a complex design phase, the project – lasting a year in total – has now reached the final stage of implementation and User Acceptance Testing (UAT).
Innovation and AI with Azure
In the era of digitization, the integration of cloud solutions and artificial intelligence is one of the main drivers of transformation of traditional working models.
Our most recent project for [Client Name Confidential] demonstrates how the adoption of cutting-edge technologies can generate innovative, efficient and sustainable solutions capable of meeting the challenges of the present and anticipating those of the future.
The context and requirements
The client exposed to SORINT.lab the need to design a modern and robust architecture for a new management platform.
The main challenge was to create a unified infrastructure that would allow:
- Effective data management: the need for a system capable of handling large volumes of data, from diverse sources and often in non-standardized formats.
- The automation of data extraction: the requirement to automate the process of extracting information from unstructured documents, minimizing manual intervention and lowering margins of error.
- An intuitive and accessible platform: the goal of developing a platform that was user-friendly, allowing users to interact with data and gain insights through dashboards and reports.
The Technological Solution
The implementation was based on an all-cloud-native architecture, taking full advantage of the potential of Microsoft Azure and its PaaS/SaaS components.
The initial draft of the project had some architectural limitations. After careful analysis, we proposed and implemented a series of targeted interventions that led to significant improvements in efficiency, security, and performance.
The process was structured into several key phases, each of which helped to build a robust and scalable solution.
The first phase involved a deep understanding of the needs and critical issues expressed by the customer. Through constant dialogue and active listening, SORINT identified strategic areas for intervention, with a focus on security and performance optimization aspects.
A Zero Trust approach was adopted, based on the principle of “never trust, always verify”: every request is authenticated and authorized, ensuring the protection of the system and data from internal and external threats
Core Architecture
- Azure Data Lake: Central data storage repository;
- Azure Data Factory: ETL tool for data;
- Azure AI Document Intelligence: AI component for intelligent data extraction;
- Power Platform and Power BI: Interfaces for data submission and visualization.
- On-Prem Data Gateway: Enables a secure bridge between local data and Microsoft cloud services, such as Power BI, PowerApps and others, allowing data to be accessed and transferred without directly exposing enterprise systems to the Internet.
Our Workflow
- Data Gathering: Users submit unstructured data via Power Platform, which provides some sort of interface to allow it to be uploaded.
- Data Upload: A Function App, written in .NET, is hired to move files to the Data Lake storage.
- Upload to Document Intelligence: Another Function App is engaged to take the files and send them for analysis to Document Intelligence.
- Processing and Tracking: The Document Intelligence component processes this input data and produces output that is reported to Data Lake. In addition, the SQL database tracks users and their activities, updated through a script that synchronizes with the B2C tenant.
- Data Processing: A Data Factory takes data from Data Lake and the SQL database, processes it via pipeline, and exports it finished in Excel format, needed for Power BI reports. The database keeps track of who uploaded each file, facilitating matching.
- Web Interface: A web interface, hosted on App Service, displays the data, accessible via authentication with Azure AD / B2C for external users.
Value Added
The cloud-native approach combined with the use of artificial intelligence has generated several significant benefits:
- Reduction of manual labour: Automation of data entry processes, eliminating repetitive tasks and reducing the margin of error.
- Accelerated processing time: Reduced processing time, enabling faster response to business needs.
- Optimized costs: Pricing model based on actual usage, significantly improving economic efficiency.
- Intuitive and user-friendly interface: Easy access and management for all stakeholders, making interaction with data faster and easier.
With optimized pipelines and intelligent integration of AI models, data is processed faster, enabling results to be obtained in virtually real time. This has led to a significant increase in productivity and a significant reduction in overall work time.
From an economic perspective, the adoption of a cloud solution has enabled the implementation of a model based on costs proportional to actual usage, improving economic sustainability and optimizing resources.
Future Perspectives
This project is a concrete example of how the integration of cloud computing and artificial intelligence can transform complex business processes into efficient and scalable solutions. The combination of Document Intelligence, PaaS and SaaS services has resulted in an innovative solution, laying the foundation for future implementations of AI-based technologies in business processes.
The platform not only meets current needs, but also paves the way for future developments and integrations, confirming the value of technological innovation in the contemporary business landscape.
Our collaboration with [Client Name Confidential] demonstrates how crucial it is to embrace innovation to meet modern challenges and make strategic decisions that promote growth.
References and further reading
https://azure.microsoft.com/en-us/products/ai-services/ai-document-intelligence
https://azure.microsoft.com/en-us/products/data-factory
https://azure.microsoft.com/en-us/products/storage/data-lake-storage