Artificial intelligence is no longer a hot topic at tech conferences; it has simply become the new norm. While many sessions at trucking industry conferences continue to center around how fleets can and should use AI to improve operational efficiency, the vendors that provide fleet technology have been announcing enhancements to existing AI solutions that many of these fleets already deploy.
One such vendor is Lytx, a longtime video safety provider that expanded in 2025 to offer a comprehensive fleet management solution via a partnership with Geotab with the launch of Lytx+. At the beginning of this year, the provider rolled out its own all-in-one fleet management solution called LytxOne, which merges video safety, telematics, maintenance, compliance and asset tracking into a single platform built natively from the ground up as a unified ecosystem.
Lytx unveiled Monday at its Protect 2026 user conference in San Diego that it is beta testing AI functionality that is designed to transform data from its all-in-one platforms, which now also includes an integration with Platform Science, into actionable insights with the Lytx AI Assistant. It assists employees in making informed decisions, while the employee still manages the workflow and owns the outcome, but the company is also introducing AI Agents in which the employee still makes the decision but hands the work over to the agent.
Lytx Chief Technology Officer Rajesh Rudraradhya showcased AI Agents to the audience Monday. He later told CCJ that customers offered excited feedback following the announcement as well as ideas for additional agents.
“I remember thinking (customers) will be freaked out by the agents, but it was to the converse, which was nice to see,” he said.
AI Agents
Lytx’s new AI Agents solution is a group of purpose-built AI workers – not generic, off-the-shelf AI – trained on a fleet’s data, designed for their workflows.
It includes Risk Analyst, an agent that triages driving events 24/7 and uses reasoning to understand behavioral context versus situational context.
“The end result is that your team only reviews the right events, not all of them,” Lytx Chief Technology Officer Rajesh Rudraradhya told the audience.
Next, is the Campaign Builder, which can assemble a training campaign for managers based on a library of events within minutes. The campaign’s design is centered on individual fleets’ data, accounting for things like route types, driver events, etc., allowing fleets to focus on their specific pain points.
The third agent is called the Fleet Strategist. It identifies patterns that are too complex for humans, correlating driver events, weather, routes, etc. to surface smarter operational decisions.
“You’re constantly balancing safety against efficiency, compliance against speed, getting it done versus getting it right,” Rudraradhya said. “Quite often, you don’t have a choice. You have to find a way to do both under real pressure with imperfect information. That’s what I call the invisible tradeoff. When you make these decisions, there’s always a tradeoff that you don’t see.”
He said this agent reveals the previously invisible tradeoffs as they’re made.
Much like teams collaborate for the best outcomes, these agents can talk to each other.
If a truck breaks down, the safety, maintenance and operations teams come together to address the situation, Rudraradhya said, noting that these agents can do the same thing, allowing people to hand off tedious, manual work to AI. The people can set the direction or parameters for the agents and let the AI do the heavy lifting.
Augmenting labor
Rudraradhya said it’s not about replacing workers.
While fleets have capable people in their operations, he said they make imperfect decisions because they don’t have the full context of their vast datasets like AI. AI Agents equip them with technology that augments their roles, allowing them to relinquish a well-defined part of their workflow for AI to manage so they can apply their experience at scale.
“Your role matters more than ever when we introduce these agents. Your knowledge, your instincts, your judgment, how you lead and motivate teams, how you make strategic decisions, how you own outcomes, no AI is going to replace that,” Rudraradhya said. “AI can handle what headcount can’t. Imagine trying to process a billion video events. AI can easily handle that sort of scale, whereas, if you ask the human team to process a billion events, it’s going to take time.”
He said Lytx has a set of pillars that guide its AI Agents, establishing a foundation of trust to ensure customers can have confidence in the AI’s ability.
“Trust is that magic ingredient that actually enables AI and humans to work with each other,” he said.
The first pillar is explainability. Rudraradhya said the system will make the AI’s logic behind each recommendation visible. The second pillar lies in the customer’s ability to control the system, allowing them to adjust, challenge, override and reject any recommendation the agents make.
“The system will never act unilaterally. It makes recommendations, and your people make the final decisions,” he told customers during his demonstration.
The third pillar is privacy by design.
Efficiency gains
Rudraradhya said he expects fleets using AI Agents will see a 50% to 80% reduction in manual work. During his presentation, he offered an example of a Lytx customer using the feature in beta in which an operations manager at a fleet outside of Dallas was able to rebalance shifts within minutes – a task that could take anywhere from hours to days when performed manually.
He said he also expects a reduction in downtime by 30% to 50% because the predictive AI models have gotten so good that they can flag mechanical failures about two weeks before they happen, turning roadside events into scheduled repairs.
Additionally, he said he expects a significant reduction in claims because the system is constantly working to prevent them.
“Every role in your company that touches fleet management will benefit from the AI Agents that we are adding into the platform today,” Rudraradhya said.