Azure-Based Lending Platform Powered by AI

About Our Client

The Client is a US-based fintech startup focused on innovative lending solutions. The company offers data-driven private credit solutions that facilitate open market-making and ensure transparent sourcing and underwriting processes.

Expert Help Needed to Plan an AI-Supported Lending Platform

The Client wanted to develop a multi-functional lending platform that would enable intuitive opportunity search and peer-to-peer (P2P) engagement for private creditors and business borrowers. The platform was to provide automated risk profiling and loan origination. The startup also wanted to implement AI-powered analytics to provide borrowers with intelligent advice on the best-fitting loan options, enhance credit risk assessment, and enable data-driven loan term optimization for lenders.

The Client sought professional assistance to create a technical design for its platform and accurately plan product development. ScienceSoft came across as a trustworthy consulting partner with decades of experience in engineering full-cycle lending software.

Lending Platform Design and Development Project Planning

ScienceSoft assembled a fintech consulting team consisting of a project manager, a financial IT consultant, a solution architect, and a UX/UI designer.

Functional design

ScienceSoft’s financial IT consultant conducted a series of interview sessions with the Client’s stakeholders to understand their business objectives, needs, and constraints and elicit the requirements for the platform. The Client had a clear platform concept in mind. Our consultant helped the company sharpen its vision for the product and converted the gathered requirements into a functional specification. Our expert also researched the preferences of the platform’s target audience to align the functional design and unique selling proposition with the end-user expectations. The platform had to appeal to both lenders (banks, non-bank lenders, etc.) and business borrowers from various industries.

The Client wanted to implement advanced lending analytics and automated borrower risk scoring, and ScienceSoft’s consultant suggested secure and cost-effective AI-supported features to enable these capabilities. The consultant prioritized platform features for launch based on their value for users and chose the functional scope for an MVP, taking into account the startup’s intended market entry strategy.

Technical design

ScienceSoft’s architect designed the optimal architecture and tech stack for the lending platform. Our architect suggested applying a hybrid architecture that combines SOA and microservices patterns. Such an approach would simplify platform governance and security while enabling agile evolution and easy scaling of granular features. With SOA's emphasis on reusable logic and microservices' support for independent feature releases, the Client would benefit from reduced engineering efforts and a shorter time to market for its solution.

Building the platform on a cloud platform would let the Client quickly roll out and scale the solution without upfront investments in dedicated hosting and development infrastructures. ScienceSoft’s architect compared the available cloud platforms in terms of the Client’s non-functional requirements and recommended Azure App Service as the best match. The platform’s high security and built-in compliance tools would ensure robust protection of sensitive lending data. Azure’s auto-scaling features, ready-made integrations, and flexible pricing models made it the best option for optimizing the platform's operating costs.

The Client wanted its platform to automatically capture the data necessary for borrower pre-qualification and credit risk scoring. The architect designed secure integrations with bank systems and credit rating platforms (by Experian, Equifax, etc.) to enable no-touch aggregation of external data feeds. Our expert also planned integrations with payment gateways for the instant processing of users’ subscription payments.

Loan underwriting on the Client’s platform required lender access to borrower data and documents. ScienceSoft’s architect suggested storing structured KYC data in Azure SQL Database and documents — in Azure Blob Storage to reduce storage costs. To enable robust lending automation, predictive, and assistive functionality, the architect chose Azure Machine Learning and Azure AI Document Intelligence services. The availability of pre-trained ML models for data processing and opportunity matching minimized the need for costly custom AI engineering and reduced the platform’s time to market.

UX and UI design

The Client wanted to ensure seamless experiences for all platform members from the onset, so ScienceSoft’s UX/UI designer joined the project at the early conceptualization stage. Our designer analyzed the needs of target user groups (borrowers, lenders, platform admins, etc.), mapped role-specific user journeys, and created UX wireframes showcasing the platform’s structure and core layouts. In particular, the Client got the blueprints for borrower pre-qualification, credit provider selection, loan application, and lending analytics screens.

The Client also received a design concept and UI mockups representing the visual style of the platform. The designer created intuitive and stylish interfaces to drive high user adoption and help the startup build brand recognition.

ScienceSoft suggested applying responsive design to ensure the platform looks sleek and works well across various desktop, mobile, and tablet devices.

Development project planning

After the Client approved the proposed functional, technical, and UX/UI designs, ScienceSoft’s project manager proceeded to plan the development project. Our PM studied the feature scope and suggested breaking down the project into three major stages:

• Developing core features for the platform’s users, including user profile management, opportunity search, loan application for borrowers, underwriting and loan origination for lenders, and account management for admins. At this stage, engineers should also develop automation modules for KYC/KYB, risk scoring, party matching, and fraud detection and set up scalable storage for lending data.
• Developing functionality for automated entry and trend-based analysis of lending data (borrower behaviors and risks, loan issuance patterns, etc.). This stage covers building APIs and setting up the planned integrations.
• Developing AI-supported features for party matching. This is also the stage where the development team implements advanced analytics components for lender, borrower, platform usage, and fraud detection analytics.

ScienceSoft’s project manager suggested applying iterative Agile methodology to speed up MVP launch and optimize the project budget. In collaboration with the consultant, our PM scoped the development tasks for each stage, determined the resources needed to complete the project, and estimated project cost and timeframes. The Client also received a project risk mitigation strategy.

A Shorter Time to Market for Innovative Lending Software

In just four weeks, the Client received a clear functional concept, a technical design, and an implementation plan for its lending platform. ScienceSoft’s advice on the optimal platform design and cost-effective tech stack helped the startup reduce project budget while ensuring high product value for the target user segments. Thanks to expert help with accurate MVP design and feature rollout plan, the startup got the opportunity to reduce time to market for the platform and speed up payback. A detailed project plan and a risk mitigation strategy delivered by ScienceSoft ensured minimized development risks.

Methodologies

Interviews, workshops, user journey mapping, functional decomposition, wireframing, prototyping.

Technologies and Tools

.NET, Angular, TypeScript, Microsoft Azure.