How Guidewheel used Malleable to give its installers their evenings back
Based on a conversation with Lawrence Han, VP of Product & Engineering at Guidewheel
Guidewheel provides real-time visibility and analytics for factory operations using non-intrusive sensors that monitor equipment's electrical activity. By detecting patterns and anomalies, Guidewheel helps factories reduce downtime, optimize productivity, and enhance energy efficiency.
To perform the install, a Guidewheel team member visits the customer's factory site to install a meter on, for example, the power cord of the machine. They configure the meter, connecting it to Guidewheel's secure cloud data repository, where the customer can use the data.
The configuration bottleneck¶
That configuration was manual. It used hardware interfaces on the meter device, the local network, and a browser to the meter's admin page, to wire the device into Guidewheel's cloud. Then, the installer used Guidewheel's administrative portal to enter the customer, location, machine, and copy-paste configuration from the meter.
The task switch from physical installation to configuration was big: taking off gloves for a laptop; finding Wi-Fi. So, installers would install many meters first and configure them in a batch, recording parameters from the meter on a phone or paper. But often, a meter would lose network connection, or information would be missing, forcing a second site visit.
Automation that gets used¶
Scripts intended to automate configuration of the off-the-shelf meters were not widely adopted by the installer team. The team members are constantly on the road, so bringing them together to change a process is not easy.
“This spread unlike anything we've seen before. We can't get it out of people's hands.”
Guidewheel asked Malleable to turn the standard process for meter configuration (in fact, a PDF of the SOP) into a runnable agentic AI workflow. It uses Guidewheel's own inputs and language, with all process quirks built-in. An AI agent configures each meter, following the workflow reliably.
One of the team members sent a link to the new Malleable workflow in a Slack message. There was no formal training, because the workflow and user interface were shaped specifically to the task. Within two weeks, the entire team had switched.
With Malleable on the team¶
Malleable published the workflow during a site visit on Feb 18th. Within days, it was doing most of the work.
A big reason for this is the simplest: Each meter took 10 minutes to configure manually, while the Malleable entry form takes seconds, then can be left to work. The delta is up to 90 minutes of configuration work after a whole day of on-site installations.
The morale boost was even bigger than the efficiency boost. Malleable removes the worst part of an installer's day: repetitive computer work when they really just want to rest. That means more energy for engaging with customers and thinking deeply about their needs.
Meter configuration is just one node of a wider workflow at Guidewheel. As Malleable customers transfer repetitive standard operating procedures to AI, they can spend more time on the fun parts of the job.
Explore Malleable