How an ecosystem of service and technology partners can help you scale your automation program

The situation: A UK-based telecommunications giant had to deal with more than a million emails a year in its global business services (GBS) unit alone. The company wanted to reduce human intervention. Long-term services partner iOPEX and partners met the challenge with an automation framework infused with intelligence. So how has their approach scaled where so many get stuck at the pilot stage?

The iOPEX framework was only possible because an ecosystem of partners worked closely together. Its application spread quickly from the initial pilot in one department to the client’s whole GBS, and it has now spread throughout the whole company.

This chimes with HFS Research findings in Exhibit 1 that when they seek automation solutions, customers refer to a far wider range of partners than simply RPA vendors.

Exhibit 1: “Automation” means different things to different types of customers

Sample: 200 Business and IT automation decision-makers, (Q4, 2021 HFS Survey)
Source: HFS Research, 2022

The journey began with the client defining a problem to solve and a partnership keen to prove its effectiveness. The telco giant regularly runs “day in the life” analyses to identify areas where the business can improve. One such analysis led the client to ask its platform services group to identify the non-value-adding work people were doing. The sheer volume of email management stood out as a vast time-sink within GBS. There was a clear problem to solve and an executive ambition to solve it.

A major customer’s need and partner funding presented a win-win for all parties

The need crystalized just as iOPEX found itself in the position to offer a free pilot. It was able to do so because Google Cloud Platform (GCP) stepped up with an offer to fund the pilot, intent on proving the capabilities of its Auto ML for natural language processing, translation, and classification. Supply met demand at just the right moment.

Like iOPEX, Blue Prism was a long-term incumbent at the telco company, so Blue Prism provided the digital workers. Combined with the application of Google AI, iOPEX could infuse intelligence into those digital workers.

The plan was to start small with a single use case to prove the approach and then scale quickly across multiple use cases iOPEX already knew existed within GBS, thanks to its long-term work with this enterprise.

Exhibit 2: iOPEX framework needs an ecosystem of partners to deliver

Source: HFS Research, 2022

The solution: A scalable framework to triage and classify emails

Work started on examples such as managed inquiries within billing and the Invoice to Pay Helpdesk within finance. iOPEX created a framework supporting the triage and classification of emails to remove the need for human intervention from multiple use cases across the GBS that could also serve the wider enterprise.

The automated process opens email mailboxes and ingests unread emails. It extracts attachments and images and converts content to text via Google’s OCR. The digital workers (bots) then automatically translated queries to English when needed. The framework then determines the content of the email and applies ML and NLP techniques such as sentiment and entity analysis to trigger the required action, such as classifying emails into various predefined categories like complaints, service requests, and inquiries.

The incident management system (ServiceNow) and ERP system (Oracle) are integrated, enabling auto-ticket creation and resolution suggestions. When the problem is resolved, details are available in the learning model, automating a cycle of improvement to auto-resolve more and more issues. Where an issue is unknown to the system, it is directed to a human for resolution, and the learning from that is fed back into the learning model.

For example, in the case of a service request, the model auto-fetches the relevant responses from cases already in the learning model, which is in turn connected to the ERP, where the data resides.

The Results: Customer satisfaction up, human intervention cut, $15 million saved

iOPEX delivered an intelligent framework to automate the classification, extraction, and detection of email content for customer service. It can be auto-scaled to meet volume and to handle additional use cases. It has since been extended to include customer queries from multiple channels, such as chatbots. Customer satisfaction scores have jumped to 92%.

In hard numbers, iOPEX says the telco giant has reduced human intervention in handling emails by 80%, cutting the time to route an inquiry to the right person from 15 minutes to less than two and delivering up to $15 million in cost and labor savings. It’s clear that the company has had to add less staff to handle increases in email volumes. But, as in most cases where automation is applied, there is one metric we aren’t seeing: How has the free time that humans are, in theory, being provided with been converted into value?

The Bottom Line: Don’t ask yourself what you should automate. Focus on the problems you need to solve.

Enterprise leaders should not have to ask themselves, “What should I automate?” They should instead focus on the problem they need to solve and then turn to their partner ecosystems to support solutions to the problem. The fact that the iOPEX solution has scaled beyond the pilot should encourage other enterprises to focus on the problem they want solving and align themselves with ecosystems of trusted solutions providers consisting of services and technology partners.

The iOPEX project proves automation can make a significant impact in large enterprises. We applaud the integration of enterprise technologies (such as ServiceNow and ERP Systems) in the solution and recognize the deep customer knowledge iOPEX was able to draw on to identify how and where to scale out beyond the pilot. But this case shows it takes more than technology and length of tenure to succeed. You need to align enterprise will, services, and technology vendor ambition to move the needle.