Zum launched Zoe on November 10, 2026, making the conversational AI assistant out there inside Zum CMX for varsity transportation groups, operators and directors. It analyzes district working information and proposes subsequent steps, whereas leaving route adjustments, driver assignments and household communications below workers management.
Zum studies that greater than 7,500 colleges throughout 19 states have relied on CMX. The launch places conversational AI inside that scholar mobility platform, with the fast intention of lowering the studies, spreadsheets and handbook cross-checking required to research late buses, protection issues and payroll anomalies.
Cross-System Evaluation for Every day Operations
A consumer can ask a plain-language query and obtain a synthesized reply, a possible trigger and a instructed subsequent step, along with underlying information for evaluate. Zoe might be opened via the Zum Assistant button or an Analyze button on a web page. It additionally provides step-by-step CMX steering from Zum’s information base.
The documented workflows cowl on-time efficiency, route protection and payroll. For a late route, Zoe can distinguish an remoted incident from a recurring sample and study potential causes equivalent to driver protection, route design or unscheduled stops. Suggestions can embody driver teaching or a route adjustment, with the supporting information out there for checking.
That evaluation attracts on CMX’s built-in information: routes, runs, telematics, ridership, payroll, driver assignments and scholar data. Connecting these information permits a missed on-time outcome to be examined alongside additional stops and paid-hour patterns with out separate exports. A payroll evaluate can due to this fact begin with uncommon paid hours and slender to particular shifts involving repeated additional stops or unexplained dwell time.
For protection, Zoe identifies routes that repeatedly want cowl drivers or are incessantly break up, and might examine a present route or run with an earlier model to summarize adjustments. Different queries embody whether or not a driver stopped at each cease, what modified on a route because the earlier week and what’s driving a day’s efficiency.
Privateness Controls and Human Oversight
The official product documentation says Zoe follows Zum’s present role-based permissions and doesn’t develop the information a consumer is permitted to see. Pupil data stays in Zum’s safe cloud, encrypted in transit and at relaxation, and isn’t supplied to AI mannequin firms for coaching. Zum doesn’t promote or share scholar information.
Solutions come from the district’s personal information fairly than the open web, and proposals embody proof that customers can examine. At launch, Zoe investigates, analyzes and recommends; it doesn’t change routes, reassign drivers or contact households. Stay help stays out there alongside the assistant when a query wants an individual or a consumer has a security concern.
District Analysis and Deliberate Actions
Tomas Beccar Varela, interim government director of transportation at San Francisco Unified College District, mentioned his district was evaluating Zoe and noticed promise in making complicated transportation information simpler to entry, interpret and analyze. He described an analysis of the capabilities, not a accomplished districtwide deployment.
Greg Massa, a transportation methods specialist at Virginia Seaside Metropolis Public Faculties, described the impact on route critiques in Zum’s technical announcement: “With Zoe, we will rapidly see what’s modified on a route, perceive how these adjustments impression efficiency and determine changes to enhance reliability.”
Zum says it’s constructing towards actions equivalent to reassigning a canopy driver, adjusting a route or following up with a household, topic to approvals and the identical privateness protections. That roadmap would transfer Zoe from investigation into execution, the place a advice can turn out to be an operational change. The approval course of and proof proven to workers then turn out to be the management level for preserving human accountability because the assistant’s function expands.
