The Company

Workfellow was founded in 2020 by a group of former Cap Gemini consultants who wanted to apply what they had learned from their collective project experience to create a knowledge mapping platform. With CEO Kustaa Kivelä at the helm, Workfellow raised a seed round of $3 million in December 2020. The round was led by OpenOcean, with participation from Icebreaker.vc and angel investors Kedar Dani and Kulpreet Singh. Employing about 25 people, Workfellow currently has 10 paying customers, including the Finnish state postal service Posti and financial services system integrator Staria.

The Technology

In creating Workfellow, the founding team drew on their experiences seeing organizations in difficulty when trying to implement complex projects without the required process maps to provide detail and guidance as to current operations. Workfellow estimates that 70% of organizations have little to no process mapping in place and are unable to manage the various data sources and activities.

Workfellow is designed to be a simple, self-service platform that is commissioned and largely operated by lines of business rather than a dedicated data science or specialist team. It is designed to provide the sort of insights that usually come from task mining and process mining platforms, without the integration overheads of the latter or the employee focus of the former.

Using what Workfellow calls its “Work API,” event logs are created from multiple applications and sent to the platform via plugins installed on team members’ computers. These record a strict and pre-defined, distinct set of activities from applications being used.

By disassembling the querystrings associated with browser-based applications, for example, one can determine the task being undertaken and customer URIs. Using this data across teams (which clients define through the Workfellow interface) it is possible to analyze the customer-centricity of tasks and processes. The Work API plugin can also monitor thick client applications by using the executable and system logs. Every application is specifically whitelisted and only activity within the specific list is recorded. Identifying that computer with a nominated, customer-defined team is as good as it gets when it comes to employee data. Workfellow does not store personal data.

The Workfellow platform’s UI allows a “Workview” graphical overview of the paths workers are using to move between applications as they perform their duties (see Figure 1). For example, looking at a customer case within a service application, the worker jumps to an ERP to check information before updating a CRM record. As the URI information is captured as part of that specific dataset by Work API, it is possible to identify specific customer cases, meaning that resolution paths can be identified and analyzed to understand their relative success. These Workviews can be filtered down by team, application, and time spent (e.g., throughput times for tasks) and volumes of data can also be reported upon within the interface.

To date, Workfellow has found a couple of core uses for the platform with early customers, most notably process discovery and process harmonization. The discovery element – the original raison d’etre for the company’s founding – is unsurprising, but Workfellow states that customers can get insights on process health using the platform in a matter of days, far faster than the alternative, manual methods previously used. Customers can then move toward harmonization, enabling best practices identified by Workfellow’s analysis to be established and reinforced within the organization.

Workfellow’s short-term development plan over the next two quarters is focused on the Work API and insight delivery. Work API will gain a studio UI which is envisaged to provide more customer control and configuration around how data is collected and fed into the platform, as well as adding more data-point-collection options. On the insight side, Workfellow intends to build a simulation feature so proposed process improvements can be run to see how they might affect defined performance KPIs. Further out, the company eyes a greater range of data being able to connect to the Work API, potentially for streaming data sources, and a move toward making the platform able to augment the human/digital process in a more orchestrated manner, with greater decision intelligence guiding these capabilities.