The Company

Established in 2020 by former Avigilon COO James Henderson and co-founder Soren Frederiksen (mindzie’s CEO and CTO, respectively), mindzie is headquartered in Dallas, Texas with employees spread across the globe. The company’s founding was inspired by Henderson’s experience with his former employer, where he led a process improvement program entirely informed by manual reporting prior to the company’s acquisition by Motorola in 2018. mindzie targets midsize and larger enterprises with tooling designed to be placed in the hands of business users, rather than data scientists or developers. The company raised a $2.3 million seed round from undisclosed investors in January 2021.

The Technology

The creation of mindzie was informed by the prior experiences of the leadership team in establishing process improvements. Doing this work department by department involved lengthy in-person interviews and manual mapping sessions with subject matter experts. The sheer length of time to enact change, coupled with the lack of real data-driven methodologies, eventually led to the development and establishment of its US-based, midsize-enterprise-focused process mining platform to simplify this work.

The goal of mindzie is to help executives move the needle on targets they need to hit over the next few quarters. So, rather than a broad-brush approach to process discovery, mindzie takes a more pragmatic approach, helping identify and improve specific problem areas flagged and prioritized by an organization’s leadership. Inevitably, this means mindzie is typically used in engagements targeting revenue and cost centers; for example, “order to cash”/”lead to close” for revenue and support tickets/procurement/accounting use cases for cost control (see Figure 1).

mindzie’s approach is to first onboard existing enterprise application data – for example, from ERP systems. However, any structured data can be analyzed, hence a range of common data connectors are provided. Once ingested and anonymized, the data is classified and clustered to normalize it through an enrichment process of adding attributes (e.g., task labeling) and performance data and removing outliers as required. mindzie suggests that a typical ramp-up time from inception to the delivery of initial results and recommendations is generally 2-4 weeks. It does not use generic indicative data to influence the recommendations, preferring instead to populate underlying models with organization-specific data. mindzie takes this approach as it believes there is too much variance from customer to customer to make generalized, pre-populated data models worthwhile.

Once this work is completed, several outputs are shown on a customizable dashboard. Primarily these are process maps that can be interrogated through mindzie’s own low-code analysis tool and output using pre-built report templates, processes generally assigned to business analysts on customer projects. Digging further, what mindzie refers to as “Variant DNA” can be reported upon, showing the various paths through established processes that have been identified from the analyzed data, with metrics for each, including relative performance. This can be used for use cases such as root cause analysis (RCA) to understand why process variants might not meet conformance standards, and it is able to pick up “stuttering” within executed processes, where steps are repeated without a clear reason why. From this, mindzie can also recommend enhancements using AI trained specifically on the ingested, normalized customer data combined with input from mindzie’s own human analysts and business analysts from the customer’s team.
Part of mindzie’s go-to-market is through independent consultants that utilize the platform within their engagements. This makes good sense as mindzie can recommend processes that it believes are suitable for automation – “automation scoring” – and plug this into a robotic process automation (RPA) queue for attention. Given the proliferation of RPA tools and the complexity of the processes that some enterprises are attempting to wrangle using RPA, having a hook into the consultancy market feels smart at the current time.

Unlike traditional process mining, mindzie has been built to be used widely across a client’s organization rather than simply employed from project to project – particularly as data from connected systems can be updated on an ongoing basis, improving AI-driven recommendations and ensuring continued adherence to process conformation. In addition, mindzie focuses on operationalizing process mining with proactive alerting and predictive AI that provide leaders the insights they need to actively continue to improve their operations. mindzie is available as a SaaS platform via Microsoft Azure. However, it also plans to be available as an on-premises application starting in early 2023 for customers that wish to continue to operate their own discrete data centers.
In summary, mindzie provides low-code, template-driven process mining tools that can be used by non-technical business users in partnership with business analysts to analyze, understand, and improve complex business process activities.