Microsoft harnesses traditional AI, Gen AI and machine-learning
Published: Sep 2025
Harnessing the Power of Technology
Highly Commended Winner
Microsoft
Photo of Houda Zeroual and Gisela Vivian, Microsoft.
Davide Musillo
Group Finance Manager
USA
Founded in 1975, Microsoft Corporation is an American multinational corporation and technology conglomerate headquartered in Redmond, Washington.
The challenge
Microsoft has experienced significant growth in revenue, billing and customers during the past decade. The level of complexity to support and provide services to its expanding base of customers has also increased and this has led to an extremely complex treasury credit and collections eco‑system with multiple tools and systems, some being third-party propriety systems along with fractured processes with heavy reliance on manual activities.
The fragmented nature of the systems and disjointed processes conflicted with the business requirements of accelerating cash velocity and has resulted in having to manage duplicate activities and processes across multiple systems, ultimately generating additional operational burden and inconsistent customer experience across lines of business (LOB).
Some of the challenges faced include a lack of systematic ability to have a 360-degree view of a customer across all its business segments. Customer master data for the same customer living in different systems for each LOB which hindered the 360-degree view of the customer.
Management of millions of inbound and outbound customer emails per year involved manual categorisation, tracking and assignment as well as fully manual responses to all correspondence received. Multiple systems needed to be configured with collections strategies with over 160 separate strategies hard coded within each system. To support these strategies, the company required 800+ templates to manage correspondence with its customers.
Multiple systems were used to manage the various collections activities including the outreach to overdue customers. This led to inconsistencies in how Microsoft provides service to its customers and results in manual workarounds and offline approvals for various items such as refunds, write-offs, holds, releases and any item that requires a level of review and approval, which all lead to inefficiencies and are open to compliance and risk gaps. Processing the volume and value of cash applications across multiple systems is complex and the percentage of auto‑posting with no touch is difficult to increase. This resulted in a lot of manual work to post payments to the correct customer account.
The solution
To keep improving cash velocity, operating efficiency and customer experience, Microsoft treasury took on some of the key challenges. It initiated the Microsoft Integrated Collections Solution Project (aka Project MICS), a two-year project in partnership with Microsoft engineering teams as well as its strategic partner Accenture.
The goal of project MICS was to adopt a customer-centric approach, simplify strategies and policies, build a technology ecosystem with automated, integrated and standardised processes. Artificial intelligence (AI) is embedded in the solution to proactively support agents by eliminating repetitive low-value tasks through a combination of traditional AI with machine learning (ML) and generative AI (Gen AI) using Copilot and Azure Open AI, which is natively integrated in its Dynamics 365 environment. An additional value-add of project MICS was building a solution that would leverage and showcase the best of Microsoft Cloud and AI technology, both internally and externally.
Best practice and innovation
Microsoft collaborated with its skilled engineering teams to leverage the capabilities of Microsoft products (Dynamics 365, Copilot, Azure Cloud services, Azure Open AI, Power BI, etc), and developed a new solution named Microsoft Integrated Collections Solutions (MICS). By harnessing both traditional AI/ML and generative AI technologies, MICS automates key activities across various workstreams:
Customer master data and credit management – consolidated and standardised master data across multiple LOBs serving as the foundational layer for AR consolidation, automated API look-up for master data consolidation and duplicate identification, contact updates workflow for customer-initiated changes and system’s prompts to review/lift accounts based on criteria.
Customer communications – automated categorisation of incoming emails, case creation, prioritisation and assignment, content summarisation and AI agent assistance in drafting responses.
Collections – simplified and standardised collection strategies and correspondence templates, worklist prioritisation for agents, automates payment reminders and overdue emails, proactively identifies invoices likely to be disputed and integrated workflows for follow-up actions.
Payments posting – enhanced the straight through processing of payments using smart cash application logic and leveraged Copilot to assist agents in customer’s payment research.
Key benefits
Cost savings.
Headcount savings.
Process efficiencies.
Return on investment.
Increased automation.
Risk mitigated.
Improved visibility.
Errors reduced.
Number of banking partners/bank accounts reduced.
Manual intervention reduced.
Increased system connectivity.
Future-proof solution.
Exceptional implementation (budget/time).
Improved key performance indicator (KPI) metrics.
“The MICS solution has resulted in saving approximately 84,000 hours annually, leading to multi-million-dollar vendor cost reductions, increased cash velocity and an improved experience for both customers and users,” explains Davide Musillo, Group Finance Manager.
“The development of Microsoft Integrated Collections Solutions brings together the best of our existing technology, using Generative AI and machine learning to automate and simplify every aspect of collections – helping us serve hundreds of thousands of customers and partners worldwide, while accelerating collections and increasing cash velocity,” said Jose Luis Marti, Senior Director.
The Adam Smith Awards are the industry benchmark for best practice and innovation in corporate treasury. The 2025 awards attracted 454 nominations. To find out more please visit treasurytoday.com/adam-smith-awards
Please enter the email that you signed up with below. If your email is
connected to a member account, we will send you a reset link.
This website uses cookies and asks for your personal data to enhance your browsing experience. We are committed to protecting your privacy and ensuring your data is handled in compliance with the General Data Protection Regulation (GDPR).
