RPA vendor WorkFusion looks to drive smarter AI

The growth of digital transformation is stimulating accelerated growth of digital work processes, and WorkFusion thinks its new product release gives it strong momentum to take advantage of this opportunity.

Robotic process automation [RPA] vendor WorkFusion is looking to capitalize on the strong momentum in the sector, with a major new product release and a favorable new analyst report.

New York City-based WorkFusion is one of the top players in the RPA space, which is booming. A new report from Zinnow on the RPA market reported that the total worldwide addressable market is $50 billion, with enterprise spend predicted to grow at a 37 per cent rate. Forrester Research, one of the analyst firms who specialize in the area, scored WorkFusion as a strong performer in their 2018 report, behind UIPath, Automation Anywhere, and Blue Prism. WorkFusion also scores very highly in a 2019 report by Everest Group, another analyst firm with a strong focus on RPA. The Everest Group report on a sub-space, Intelligent Document Processing [IDP], defined as a software solution that captures data from documents and categorizes and extracts relevant data for AI processing, ranked WorkFusion in the leader section, with the highest score in vision and capability.

“Everest Group ranked us high in the IDP leader segment, and they see our use of intelligence as a key differentiator for us,” said Sam Fahmy, WorkFusion’s CMO. “WorkFusion is the only platform built on AI as opposed to being built on algorithms. We believe true automation is not simply moving things from point A to point B. That’s just wide-ranging automated copy-and-paste. To truly automate moving data isn’t the ultimate goal – labelling and quantifying it is. That reading of unstructured data – and making decisions with it – that’s what we are really good at. Intelligence is understanding what’s in the data, and what to do with it.”

Founded in 2010, WorkFusion is primarily, although not exclusively, focused on the enterprise.

“In Canada, three of our most successful customers are very different,” Fahmy said. “BMO, a large bank, is very traditional. Polaris, a logistics company, is not a traditional enterprise. They use us to automate things like customs processing. With them we replaced manual tasks, and the people who used to do them now do more meaningful work at the company. With Canadian Tire, they use our technology to automate different processes, and they often know more about what our tech can do than we do.”

WorkFusion’s initial sales motion was direct, reflecting the initial use case specificity of automation, and much of their business is still direct.

“We do have a strong channel, however, which does both consulting and implementations,” Fahmy said. “We serve all verticals, but from outbound flow we are focused on six or seven, and partner with large systems integrators and others who do as well.” They have over 50 partners in their channel program, although that also includes technology partners.

The new Intelligent Automation Cloud, WorkFusion’s new AI-powered platform is designed to make the company’s technology usable by a broader audience. WorkFusion’s proprietary low-code machine learning [ML] logic allows non-technical business users to train an automation bot faster than they can train people. This removes what WorkFusion calls the greatest impediment to intelligent automation, the lengthy process of labeling and analyzing massive amounts of data and the continuous coding of bots by ML engineers.

“This new platform allows for the democratization of machine learning, to make it available to the masses – not just data scientists,” Fahmy said. “We are a leader in that space – not compared to Google, Amazon or Microsoft, but to the specialized vendors in our area.”

Also new with this release is WorkFusion’s 1–6–12 go-live program, which offers a quick time to value.

“One of the biggest pitfalls in automation is the long time to value cycle,” Fahmy said. The 1-6-12 program involves a 1-day feasibility study, which can identify good use cases to be automated, which is something that could have taken weeks before. In 6 weeks, we provide a fully working proof of concept, and by 12 weeks, will be fully in production.”