Zeitworks: MVL’s focus on machine learning and automation

By Jay Bartot (CTO) and Keith Rosema (Partner) of Madrona Venture Labs

At Madrona Venture Labs (MVL), we look for massive opportunities and incredible A-teams to be the foundation of our companies. Zeitworks founders Ryan Windham, Ben Elowitz, Matthew Holloway, Dan Henrick, and Brian Mitchell are that team. Incubated inside of Madrona Venture Labs, Zeitworks announced this week they raised a $4.5 million seed investment led by Madrona Venture Group with participation from Jazz Venture Partners and ​Spencer Rascoff​, co-founder of Zillow and dot.LA. Zeitworks’ mission is to maximize human potential at work, helping companies improve the efficiency of their operations, exposing bottlenecks, and driving improvements in real-time.

Zeitworks represents another major addition to MVL’s portfolio of vertical machine learning (ML) companies. Leveraging the growing efficacy and applicability of ML technologies to business optimization applications, Zeitworks automatically finds patterns in repetitive back-office processes (e.g. bank loans or insurance claims processing). The machine learning algorithms are fed with passively collected data created by users’ actions as they go about executing their daily work. The combination of this data and technology creates a treasure trove of actionable information for workers and managers to measure and optimize the efficiency of their businesses. 
 
Following our usual rigorous idea validation process, we tapped into our wide network for potential customers who could provide feedback on the idea. Many conversations with organizations with teams performing repetitive data processing tasks told us that they would value a solution that could give them visibility into the efficiency of their existing processes and also help find ways to be more labor-saving and cost-efficient.

As with other data-heavy ML companies incubated at MVL, we were eager to collect data to validate our hypothesis that ML technology could be instrumental in providing an automated solution. We knew from experience that without the right data the machine learning algorithms would not be likely to yield much value. Getting sample data would be an important step towards the next stage of idea validation.

Capitalizing on the encouraging response to the idea along with continued support from Madrona Venture Group, we began developing a data-driven, automated, business process analysis and modeling system, later named ‘Zeitworks.’ Key to this effort was assembling an experienced team representing disciplines including business and customer development, data science and ML, Windows client engineering, design, and product management. Team members from MVL included Ben Elowitz and the two of us (Jay Bartot and Keith Rosema). Matthew Holloway, Dan Hendrick, and Brian Mitchell were critical hires and co-founders, forming the basis of the data science, product, and design teams. 

We quickly deployed a prototype of the technology, testing it with an early pilot customer.  Initial analytical experiments with the data were promising! Each day, more and more data provided increasing insight into the signals inherent in the data, and the ML methodologies that were most effective in exposing the patterns we sought. Soon we began delivering our first set of reports, starting an iterative cycle of customer feedback in the search for product-market fit (PMF). 

In addition to building the Zeitworks core team, MVL worked hard to give the new company every possible advantage by filling in many of the gaps that usually exist for nascent teams. We provided security and privacy expertise, engaged an outsourced code development team for scale and worked with cloud providers to secure discounts and consulting time--all to get the product off to the strongest possible start. We worked with a patent attorney to lock down key intellectual property. We also put into place the core company infrastructure--from source code control and testing hardware to health plans and financial tools--that would help everything run smoothly. Finally, it wouldn’t be a startup without fun perks, so we included the team in our co-working space at Create33 and for six months they worked alongside our other founders and startups. We find this is one way we continue to learn and grow while making ourselves readily available to the teams we are incubating.

Saving the best for last, when we met Ryan Windham, former CEO of AppXView and Cedexis, a Madrona Venture Group company, we knew immediately that Ryan was the CEO for Zeitworks. After taking the helm, he and the team secured a seed round of venture capital financing and moved to spin the company out of Madrona Venture Labs. Within a few months, Zeitworks was formed and funded and ready to leave the nest. We’re proud of what Zeitworks has accomplished so far, and we’re excited for their journey ahead.

At MVL, we’ve just begun to uncover the many opportunities to develop companies and products that help enterprises leverage the value of their data with ML/AI technologies. Ready to become a founder? Come work with us!

Start a company with us

Recent stories

View more stories

Let’s start a company together

We are with our founders from day one, for the long run.

Start a company with us