How Virtual Blue Is Turning AI Hype Into Hard Business Results

For all the noise around AI, a lot of businesses still operate with the same old friction. Invoices arrive in different formats. Referral forms turn up incomplete. Orders sit in inboxes. Staff spend hours moving data between systems, checking fields, fixing errors, and doing work that has to be done but adds nothing to growth. This is the space Virtual Blue works in. The Auckland-based company operates across New Zealand and Australia, with more than 100 implementations, and its work is focused on automating repetitive processes, improving data handling, and supporting clients once those systems are live.

Founder Sharyn Catt describes the company’s focus as “using intelligent automation to transform businesses and help them to excel.” Melanie Visser, Director of Operations, adds that their work is about helping businesses “be better at what they do, so they can focus on what really matters to them.” Catt also gives a simple way to understand the difference between automation and intelligent automation. “Traditional robotic process automation is about giving a robot a set of rules and instructions to follow and it steps through a process in the same way a person would.” Intelligent automation adds AI on top, so unstructured information gets read and turned into a format the robot can use. “The easiest way I’ve always explained it to customers,” she says, “is that if you think of the robot as the arms and legs, and the AI as the brain.” It is a useful distinction, because much of the work Virtual Blue takes care of starts with time-consuming manual processes, such as supplier invoices, referral forms, or customer requests, before moving into a structured automated workflow.

Virtual Blue’s work with Allevia Radiology (Mercy Radiology) is a powerful example of how the company’s work is applied. Allevia Radiology had been processing their ACC invoicing and receipting manually. Staff had to identify the correct payer, submit invoices through online portals, and check patient details. The process was time-intensive and prone to error, and because of that, it was only done once a week. Virtual Blue introduced a digital worker called Matilda, which now handles 98 percent of invoices automatically, completes work in two hours rather than eight to ten, and helped Allevia move from weekly to daily invoicing, which enhanced cash flow by $200,000 per month.

The Allevia example also reflects where Virtual Blue often starts with a client. Catt says finance processes are a common entry point because they are regular, rules-based, and easy to measure. “The most commonly requested processes are accounts payable (AP) process, accounts receivable, and bank reconciliation,” she says. Those processes exist in every organisation, and the time before and after automation is usually easy to compare. In the interview, Catt and Visser describe benefits in two groups, financial and non-financial. Financial benefits include increased revenue, cost avoidance, reduced operating cost, and fewer hidden costs around systems and staffing. Non-financial benefits include efficiency, resilience, compliance, customer experience, employee experience, and sustainability. Virtual Blue advises clients to start tracking measurable benefits and build from there. The Warehouse Group shows how those ideas scale inside a large organisation. The retailer operates 218 stores across The Warehouse, Warehouse Stationery, and Noel Leeming. Its automation programme supports 16 business functions and more than 80 processes, gives back more than 40,600 hours, and delivers an estimated $15.3 million in annual business value. One Noel Leeming pricing process reads competitor pricing across nearly 4,000 SKUs, applies buyer-defined guardrails and rules, can calculate more than 1,100 promotional price changes a day, and pushes the updates to shelf-edge labels before stores open. That process removes around 6,000 hours of manual effort each year and delivers an estimated $2.5 million in net margin impact. Another automation gives the group daily control over pallet movements across its distribution network, with the case study stating savings of $860,000 per year.

Jeremy Dean, Technology Team Lead for Intelligent Automation at The Warehouse Group, highlights the impact of these systems. “Some processes must happen in a relatively short period of time and involve a large volume of transactions,” he says. “No matter how many people you assign to these processes, you cannot replicate the scale and the pace of the digital workers.”

Cambridge Clothing shows a different kind of result, where automation affects both internal operations and customer-facing work. The New Zealand menswear business had already embraced automation in finance, but one of the more useful applications sat in alteration requests. Before automation, a late Friday order had to wait until Monday for manual handling, which slowed the process and reduced visibility. Virtual Blue built a web form workflow and linked it to a digital worker called Bunny Robot. Bunny checks for new alteration orders four times a day, seven days a week, saves them to a network drive, creates the sales order in the ERP system, and keeps the process moving through the warehouse. Orders are now turned around in three days, which is 57 percent faster than the manual process, and customer satisfaction has risen to 92 percent. Cambridge Clothing CFO Graham Bass says, “Intelligent automation has been a game-changer for us. It covers at least five regular daily routines.”

Tropex and Tara Exports provide another clear example, this time in export and supply chain work. The two businesses were dealing with manual, paper-heavy processes across supplier invoices, freight invoices, and shipping-related records. Virtual Blue built a solution combining Angelica AI with a digital worker named Karen Norman. Karen retrieves invoices from a mailbox, routes them to Angelica for classification and data extraction, validates them against other systems, updates internal records, and flags discrepancies for manual review. After the first invoice matching rollout, the process expanded into freight invoice handling and the automatic filing of supplier emails. The case study says the result was more than 500 hours returned to the business each year, digital invoices that are searchable, and live updates that improve team transparency. Tropex director Joe McLeod says, “It’s not just about efficiency; it’s about future-proofing our operations.”
Implementing new solutions is only part of the journey. Virtual Blue provides a holistic, ongoing support function that ensures automations run seamlessly and reliably. “This gives our clients complete peace of mind, knowing they have a trusted expert partner managing and optimising their automation environment,” Visser says. “We really take pride in the relationship we have with our customers.” The team monitors automations, responds when errors appear, and handles issues created by changes such as browser updates or system upgrades. “Our responsibility is to provide that gold standard support and make sure they don’t have to worry about their automations”.

Virtual Blue applies some of the same logic internally. Visser says the company uses its own automations to check services, gather support reporting, and alert the team if a process does not complete on time. “We actually automate the little things,” she says, before describing service checks, support reporting, and automated alerts. Later she adds, “The way we design our processes is we get the robots to talk to us.” That approach matters because it shows the company is not only building workflows for clients, but also maintaining its own systems through the same kind of operational discipline it describes to customers.

There is also a consistent pattern in how adoption develops inside client organisations. Catt and Visser both describe a degree of nervousness at the start, especially when teams first hear about automation or agents. Visser says that often changes once a team sees a process working. “As soon as we’ve implemented one process with that team, they are the ones that come back to us and say, ‘Hey, we’ve got another idea for the robot or for the agent.’” As the terminology has shifted, Virtual Blue has also changed some of the language it uses. Catt says the company now talks more about agents and describes Angelica AI as its own product layer. She traces Angelica AI back to an earlier product they created called Smart Capture, which focused purely on optical character recognition (OCR). She says it grew into a broader product that handles customer interactions, document extraction, and workflow support. In one recent example, she says the company had gone live with a customer-facing implementation, where Angelica AI connected to a CRM through APIs and used RPA for the parts where API capability was missing.

Across the range of Virtual Blue case studies, regardless of whether it involves RPA, AI, or agents, or a combination of the above, it is clear that the most useful automation is not necessarily a massive moonshot solution. It starts with one time-consuming process, one obvious bottleneck, one job people are tired of doing manually. “Start small,” Catt says. “Bring people along on the journey.” The case studies show what happens next. Time comes back, errors drop, cash flow improves, teams stop worrying about the technology and start asking where else it should go.