Rebooting internal app support at Microsoft with Dynamics 365

May 6, 2021   |  

Internal app support for Microsoft Finance had to be simplified.

The experience employees, vendors, and partners had each time they sought help with one of the company’s many internal finance processes was too complex and fragmented. App users continually requested advanced support and custom reports.

That’s where Rajiv Maheshwari and Kris Jolma come in.

They developed Microsoft’s new Unifinance platform, a retooled internal app support experience that rolled out across the Finance organization starting in 2019.


Teams photo with Bah, Vemula, Subramaniam, Bhamidipati, Kamdar, and Lanka.
Conrad Wong (top left), Mamadou Bah, Ramchander Vemula, Kavitha Subramaniam, Akhila Bhamidipati (bottom right), Dhwani Kamdar, Sekhar Lanka, and Anusha Tondapu (not pictured), helped develop and manage the Microsoft Finance internal app support platform, a retooled experience they call Unifinance. (Photo by Aleenah Ansari | Inside Track)

Their help was needed in a big way.

At Microsoft, thousands of employees need to make everyday purchases, turn in expenses, report time off, and view paystubs. Vendors deliver and check on their invoices, salespeople track contracts, and strategic planners use finance data for what-if analysis.

A tremendous amount of activity is happening all the time.

The company was using multiple solutions to fulfill financial and regulatory functions like accounting, payroll, purchasing, revenue tracking, and compliance. To get help, employees, suppliers, and partners had to navigate multiple processes, from online forms to email aliases.

This used to add up to more than a million support tickets each year. Maintaining and adding features to these separate systems became a full-time job for six engineers.

All that changed with Unifinance, Microsoft Finance’s internal app support platform built with Microsoft Dynamics 365 and the Microsoft Power Platform.

“Our Finance user support tools were scattered and not connected,” says Rajiv Maheshwari, a principal software engineer manager for Finance Engineering at Microsoft. “They were created over the years by different groups, and it took the team’s entire capacity to maintain them. We could barely do anything beyond keeping the lights on.”

For example, two separate teams at Microsoft handle payroll, one for global and one for the US. Each team had different ways of working with the helpdesk tools, collecting data, and gathering customer feedback.

“To support our users, we had more than 300 email aliases and five different customer relationship management (CRM) helpdesk ticketing apps,” says Kris Jolma, group global process manager in the company’s Microsoft Finance division.

Digging into the problem, Jolma soon discovered that none of the platforms were connected to each other, limiting the ability to transfer tickets and gain cross-department insights.

Jolma began asking if it was possible to connect and scale these systems. “That’s where I saw the challenge and opportunity,” she says. “Was there a technology that could help us?”

[Find out how Microsoft uses Dynamics 365 and machine learning to drive efficiencies in finance. Discover how AI-powered chatbots cut the number of support tickets by 20 percent.]

Streamlining common processes with a unified helpdesk platform

Jolma and her team began collaborating with Finance Engineering to come up with a vision for a new architecture. They designed a connected system with a unified helpdesk platform that would modernize app maintenance and support, and provide a more consistent user experience.

Once the goal was clear, the team was able to quickly build the baseline solution.

We realized that if we could build up Unifinance as a platform, we would have reusable enterprise capabilities that other teams could benefit from too. That’s the beauty of building a platform versus a single solution.

– Luciana Siciliano, Microsoft Finance director

The cross-company team decided to use Microsoft Dynamics 365 as a technical foundation, adding customer feedback analysis through Microsoft Azure Cognitive Services and Microsoft Azure Machine Learning to help improve overall customer satisfaction (CSAT). They knew this solution could offer users a set of modules using a robust common architecture that would work across multiple Finance teams.

Out of the active Microsoft Dynamics 365 helpdesk platforms, the team chose one of the most robust installations, known as One Finance, as a foundation. It was feature-rich and could accommodate most of the user requirements. They renamed the solution Unifinance to better fit its vision of unifying support across multiple business groups.

“We realized that if we could build up Unifinance, we would have reusable enterprise capabilities that other teams could benefit from too,” says Luciana Siciliano, a director in Microsoft Finance who collaborated with Jolma to sponsor the project from the business side. “That’s the beauty of building a platform versus a single solution.”

Hunt poses for a photo.
Emilie Hunt is an engineering director whose team is driving much of the Unifinance project. (Photo by Emilie Hunt)

The work fits nicely into a larger story that’s unfolding across the company, says Emilie Hunt, an engineering director whose team is driving much of the Unifinance project.

“The need for a unified approach fits into a larger digital transformation that’s underway across the company,” Hunt says. “This unified helpdesk platform aligns with our vision to lead the expansion of ‘touchless’ operations across the company.”

Demand for Unifinance increases with successful rollouts

Three months after getting started, the Finance team turned the first complete business process live on the new unified helpdesk platform. The team brought on each group carefully while incorporating learnings from the previous rollouts.

“After we added one program, we were continuously delivering,” says Ramchander Vemula, a senior software engineer on the Microsoft Finance team. “We’ve been giving training to other teams to familiarize them with the platform, collecting feedback, and adjusting the solution so all teams can use it to meet their needs.”

Jolma took the lead on communicating with the various stakeholders about this opportunity.

“After we got leadership buy-in, we expanded our discussions to the other orgs within our team,” she says. “Change is scary, and every team was pretty comfortable with what they had.”

Jolma began holding weekly “office hours” calls for team members to check in or share their feedback. The team provided demos of the system and how it would work. This is an area that Jolma feels could have gone better.

“We could have given more examples of what the change was going to look like for individual teams,” she says. “Something that would have helped early on would have been to do more hands-on demos with the business groups.”

But as the benefits of the unified helpdesk platform became obvious, more and more teams wanted to come on board.

“We were initially planning to onboard only nine programs,” Vemula says. “Then we found out that teams wanted to migrate over 86 processes from Outlook distribution lists to this platform.”

The time to deploy new scenarios for finance app users has already been cut by 50 percent due to the efficiencies of the internal app support platform.

With the new internal app support platform, 120 mailbox aliases that the One Finance team was handling have been combined into a single unified helpdesk experience for users. Now, each support agent has a maximum of three mailboxes to monitor.

Better customer support and timely analytics

Microsoft Royalties and Content Ops (RCO) is a centralized service organization that provides business solutions for the management of global royalty and revenue sharing contracts. RCO internal app support, provided through a system called Roycare, was one of the first to begin using Unifinance.

Unifinance helps RCO increase efficiencies and lower costs, providing fast service to global business partners and payees while delivering prompt and accurate payments.

Another key user is Microsoft Procurement. This team launched a program called Employee Device Management to help employees have a positive experience with their devices and be more productive and engaged.

Twelve pilot regions were selected and onboarded to Unifinance. The platform now helps them reduce operations costs around manual processes with a more mature CRM featuring artificial intelligence, machine learning, and automation.

Now additional capabilities like a Finance Digital Assistant (FDA) chatbot are seamlessly integrated with the Unifinance platform. This allows multiple Microsoft groups to use the FDA, like ProcureWeb and Travel.

This chatbot is a great success. Today, it solves 60 percent of Finance-related support cases by providing self-service information to customers, preventing the need to open a support ticket.

If a user still has to engage with an agent after interacting with the digital assistant, they can do it directly from this window. They don’t need to know who they should contact or how to submit a ticket.

Collecting user feedback is also faster now, with the team’s implementation of Outlook Action Cards. These are appended to every email exchange with a support agent and sent at the close of the ticket.

“This way, we collect both continuous feedback and information on the overall experience,” Vemula says. “The user doesn’t have to navigate anywhere else—it’s right within the Outlook message. We’ve already seen a four-fold increase in feedback.”

ProcureWeb and the Finance Digital Assistant screenshot.
ProcureWeb and the Finance Digital Assistant chatbot are integrated with Microsoft’s finance app helpdesk solution, Unifinance.

Unifinance also provides a single portal combining past and current data.

This data is used to provide insights to Finance Engineering, improving processes and overall customer satisfaction. This helps reveal patterns and commonly occurring scenarios so teams can quickly address issues and streamline business processes.

Valuable information and insights are now more available across the organization, from a Predict Ticket dashboard to one on user feedback analysis. In fact, several teams have started to create tickets for each other using the new system.

Next steps for Unifinance

As Finance teams continue onboarding to the Unifinance platform, Finance Engineering is deploying additional artificial intelligence and machine learning capabilities.

“We have an innovation mindset,” says Conrad Wong, a principal software engineer for the Finance team. “We want to implement other scenarios that can save agents time and improve processes that can then be rolled out across all services.”

All that learning is making a difference.

“Every time the machine learning models make a mistake, we have a process in place for the agents to fix it,” Wong says. “We feed that back into our training process, creating a continuous improvement cycle. Dashboards also help us keep track of the models’ performance over time.”

Dhwani Kamdar is a senior program manager who manages the Unifinance portfolio on a day-to-day basis.

“The primary technical goal for the next few months is to advance the system’s machine learning, automation, and data-driven capabilities,” Kamdar says. “We’ll do this while also onboarding the remaining teams on the platform.”

The team plans to gain faster insights on tickets by automating the recognition of user intent, says Jovalene Teo, senior program manager on Finance’s Center of Innovation team.

This can be part of a bigger picture of a customer support platform across the company. With the efficiencies gained, the whole package of what this platform brings is invaluable.

– Kris Jolma, Microsoft group global process manager

“Text analytics will be the next major focus,” Teo says. “It will let the system dynamically categorize tickets, drive deeper business insights, and have more chatbot automation, to further reduce operation costs and ticket volume. The unification of support systems facilitated the collection of big data that enabled these business benefits.”

This will keep reducing the volume of support tickets and help agents quickly find the right information when they do need to address a case. In fact, the learnings and efficiency gains derived from this project have the potential to transform the way other teams across Microsoft do internal app support.

“This can be part of a bigger picture of a customer support platform across the company,” Jolma says. “With the number of aliases decommissioned and the efficiencies gained, the whole package of what this platform brings is invaluable.”

The Unifinance helpdesk has helped improve user experience by delivering a multi-channel experience, Hunt says.

“Agents are empowered to solve cases faster and avoid escalations by having the right resources and data at exactly the point of need,” she says.

The next phase in this transformation is to gain deeper insights into the operations data to proactively address issues and avoid the need to raise helpdesk tickets in the first place.

“It will enable our engineers to rapidly deliver new capabilities and AI/ML-powered experiences to all of our customers with increased business impact and efficiency,” says Jane Zhang, a principal director of engineering whose team did much of the work on the project.

Zhang says the project would not have been possible without all the different teams that pitched in to help. “Unifinance showcases the cross-group partnerships we’ve developed and co-developed with product teams,” she says.

Find out how Microsoft uses Dynamics 365 and machine learning to drive efficiencies in finance.

Discover how AI-powered chatbots cut the number of support tickets by more than 20 percent.

Tags: , ,