Cisco + Celonis

As providers of industry leading AI-powered solutions, Cisco is keen to use agentic automation to optimize their core operational processes. But before they can effectively deploy AI agents, they first need to understand how their processes run and where they can be improved. Celonis is currently being deployed in a number of commercial operations use cases, with a view to supporting Cisco’s future vision of humans and AI working in harmony.

Context: Optimizing operations with AI agents

With a commitment to creating a more connected and inclusive future for all, worldwide technology leader Cisco is an early adopter of AI. After a phase of AI experimentation, and empowering employees to use the technology in their day-to-day work so they become AI-literate, the company is starting to use agentic automation to optimize its core operational processes.

As Prasad Varahabhatla, Senior Director, Process Intelligence & Automation explains, “We have a charter of making sure that AI is adopted at scale across Cisco. We can now take advantage of all the basic foundational work we have done by getting AI into our processes.”

Challenge: Identifying process gaps

Before AI can be used to automate processes and address inefficiencies, the business first needs to understand how those processes currently run and where the gaps are.

“Cisco is a 40-year-old company, and it's my belief that we have actually got a lot of our processes fairly well automated within our enterprise systems,” says Varahabhatla. “However, given our scale, even small gaps in a process have a large material impact.”

Varahabhatla’s team turned to Celonis to find out exactly where those process issues are, so they can systematically resolve them, either by making changes in enterprise systems or by automating workflows with AI agents to complement the human workforce.

Solution: Streamlining commerce operations

Cisco began by identifying a number of use cases related to the ordering process, which includes everything from order management to invoicing and sits within commerce operations. The goal is to make booking an order as smooth as possible for Cisco customers and partners so orders get booked faster. Commerce operations is an especially complex area as Cisco sell software, hardware, and services, and each one has a very different order flow.

One initial use case was around orders that had been booked but not invoiced. By avoiding invoicing delays, Cisco can improve working capital by getting cash faster. But more importantly there’s a revenue capture opportunity because subscriptions and service contracts can start on time.

Cisco used the Celonis Process Intelligence Platform to gain full visibility into the ordering process from the time an order is booked to the time the invoicing happens, to find out what’s causing delays. Multiple root causes were identified and Celonis enabled the team to prioritize which of these to tackle first to get the greatest value in the shortest amount of time.

Crucially, Cisco is now aiming to reduce the timeline for following up on orders that haven’t been invoiced from 90 days after booking to just 30 days, which is expected to accelerate revenue cycles. With Celonis, they will also be tracking the revenue capture opportunity of delayed invoices as a metric to assess the impact of process improvements.

A second commerce operations use case relates to the perfect order. A perfect order is one where a customer or partner doesn’t need to contact Cisco until the point the order is booked. Perfect order metrics are currently around 85% but, with the help of Celonis, Cisco aim to increase this to 90% in the next six months. As Varahabhatla explains, “We need to identify the scenarios where a customer or partner seeks help so that we can remove whatever is stopping them from getting to a perfect ordering experience.”

Vision: Humans and AI working together

The intention is to expand Celonis deployment beyond these initial ordering process use cases, and use the process understanding it delivers, to achieve Cisco’s vision for a future where humans and agents work in harmony.

Varahabhatla believes that while some processes will be fully automated by AI agents, there will always be a need for humans in the loop. He says, “Sometimes you want a human to make a decision. So how do you make sure that AI has done all the legwork and can provide the person all the information that they need, as well as the conclusion that the AI has come to?”

He also emphasizes the importance of human relationships, believing that AI will make Cisco employees who work with partners more efficient in managing and maintaining those partnerships. Similarly, it will make employees who work with customers more informed, so they can deliver the best possible service. But before any of this is possible with AI, businesses must first understand their processes.