Turn operational data into action with Samsara

Most AI is built for knowledge work: drafting emails, summarising documents, and answering questions. AI for physical operations requires a higher level of context and intelligence to be effective. Work happens in truck cabs, on job sites, in maintenance bays and across operations centres, and teams in those places need the right fleet or operational context to be effective.

To make those decisions at the speed at which their people and assets already move, teams need more than the visibility that a general AI tool might provide. They need AI grounded in the operation itself, with the context and workflows to help people act. That’s where Samsara comes in.

That’s where Samsara AI comes in. Using the data already in Samsara, teams can handle recurring operational work, from guiding a driver at the moment it matters to surfacing a maintenance issue to delivering a daily briefing without someone pulling the report by hand.

Watch Video

From visibility to action

Physical operations were always generating data, but teams lacked a connected view of what was happening across vehicles, sites, and people. Samsara set out to change that by connecting telematics, cameras, and sensors to make the operation visible.

But visibility is only the beginning. The next challenge is turning signals into decisions and decisions into work.

A safety team may see a risky driving pattern only after an incident. A maintenance team may find a fault code after a vehicle is already out of service. An operations manager may have all the information needed for a report, but still spend hours pulling it together by hand.

For more than a decade, Samsara has built the data foundation for physical operations: connected hardware, software, and integrations that bring telematics, video, diagnostics, driver history, Drivers' hours, DVIRs, weather, and more into one operational picture.

That foundation includes more than raw signals. It includes the context around them: what happened, who needs to respond, which workflows follow and what decisions move the operation forward.

With Agent Studio now generally available, teams can configure, test and deploy agents that turn that context into action – whether that means guiding a driver, prioritising a repair, preparing a report, or answering a question.

Off the back of Samsara MCP (generally available last week), Samsara now provides two ways to put AI to work: build and deploy agents in Samsara, or bring trusted Samsara data into another AI environment.

Teams can interact with Samsara AI in the way that fits their work:

• Samsara Assistant helps teams ask questions, find insight and move from an answer to action.

• Pre-built AI capabilities and workflows help teams address common safety, maintenance, and operations tasks without starting from scratch.

• Agent Studio, now generally available, gives teams a practical way to move from an idea to a working agent: start with 25+ pre-built templates, or build a custom chat or voice agent, then test, govern, and push live in minutes. 

• Samsara MCP extends governed Samsara data into the AI tools teams already use when a workflow spans Samsara and other business systems, all while adhering to their permissions in Samsara. 

The experience is flexible, but the goal is consistent: help the people closest to the work spend less time chasing information and more time improving the operation.

What Samsara AI looks like in practice

The value of AI becomes clear when it’s connected to the work people do every day. 

Turn safety data into coaching that drivers actually get

A safety manager can see the data, but a personalised coaching conversation with every driver, every week, takes too much time for many to complete. Even when working with drivers is critical, that work often loses out to whatever is most urgent that day.

With agents and Samsara, a safety manager can quickly create a personalised video recap for each driver from their own safety events and dash cam footage, then schedule it for automatic delivery every week – without building each recap by hand. The agent can further provide insight into what drivers need the most attention and triage an action plan for coaching for each period.

The result is coaching coverage that scales from one admin and driver to many through the support of a simple agentic workflow.

Jordan Carriers, a 900-vehicle trucking fleet with about 800 drivers, cut crashes 30% after rolling out AI-supported coaching capabilities.84% fewer severe speeding events and 79% fewer mobile-usage events resulted along the way. 'We used to have, on average, 140 drivers in coaching every day. Since implementing Samsara, it's only about 10,' said Otis Anderson, Fleet Safety and Compliance at Jordan Carriers. He now manages coaching for nearly 1,000 drivers and does about 98% of it himself. ‘Happier drivers are safer drivers,’ Anderson said.

Turn a full year of work into a five-minute report

Monthly inspection reporting means pulling asset and inspection data tab by tab, cross-checking for missing assets and expired tags, and compiling it by hand. Some back-office teams dedicate a full-time employee to this work.

Thanks to a custom agent in Samsara, it’s now possible to pull inspection and asset data directly from the platform, compile per-division reports, flag missing assets and expired registrations, and export everything with one click. Customers can customise the template for their operation and schedule the report to run daily, weekly, or monthly.

That AI tooling frees the person who used to manually populate the report to focus on higher-value, mission-critical objectives.

Grand Isle Shipyard, a field services company operating since 1948, used to dedicate significant resources to this exact work. 'We were spending more than six figures a year on reporting and data compilation, work that's now fully automated,' said Derek Champagne, VP of Corporate Security at Grand Isle Shipyard. 'The real benefit isn't just efficiency. It's the ability to focus our people on solving bigger problems, driving innovation, and creating value that simply wasn't possible before.' What once took days of tab-by-tab clicking across 600 to 700-plus inspections a month now takes about five minutes per division.

Start every shift with a clean briefing instead of a scavenger hunt

Before every shift starts, many operations require vehicle status, daily walkaround check compliance and open issues to be assembled into a single place for reporting.

With Samsara, a scheduled start-of-day briefing agent can automatically pull daily walkaround check, vehicle status and asset data.  It can fine-tune the reporting to a single yard or the whole fleet, as a standing run instead of a manual task. It can deliver the report through multiple channels, to multiple stakeholders, and in multiple languages based on the unique needs of the operation.

Then, as soon as drivers get into the cab, they automatically receive a daily briefing. That gives teams a consistent morning ritual instead of a recurring scramble, and it catches vehicle and compliance issues earlier in the day.

'Start-of-day briefings from Samsara AI help us get set at the start of our shifts, and two-way audio is great to tackle drowsiness in the cab. It all makes the focus on helping us get better – not picking us apart or making us feel inadequate,' said Tiny Kirby, a driver at Clyde Companies.

Automate dispatch questions

Drivers run into parking, weigh-station, and policy questions on the road, often outside business hours. Without an answer close to hand, that means a call to dispatch or driver support – or a guess.

With Samsara, a driver support workflow allows drivers to communicate with AI directly from the cab and instantly get answers. Organisations can customise the workflow by uploading their own handbook and policies, and the agent considers real-time drivers' hours and location data to provide a thoughtful answer to the driver. Escalation to a real human is also possible if the driver needs deeper support.

Drivers get an answer in seconds instead of minutes on hold, and the informal in-cab mentoring that used to fall on veteran drivers is backed up by something available around the clock.

'One of our drivers talks to the in-cab AI all day to ask questions about rules, regulations, and personal conveyance,' said Shawn Walker, Safety Claims Manager at ABCO Transportation. 'It has become a hands-free guide that instantly clarifies the things he didn't quite understand before.'

Close the gap between who's driving and who's logged in

Unassigned driving, when a vehicle moves but no driver is logged in, breaks safety scoring and coaching for that trip, and it's a common source of insurance and litigation exposure when no one can say who was behind the wheel.

With Samsara, an AI workflow detects unassigned driving, verifies identity through facial recognition, name matching, and audio, and automatically calls the driver to confirm and assign them to the vehicle.

StreamTech Industrial, an industrial fleet operator, saved 500 hours of unassigned drive time after deploying their driver ID agent over a two-month period. 'Drivers used to skip logging into their assigned vehicle, creating unassigned drive time,' said Jarrett Daggs, DOT Manager at StreamTech Industrial. 'Now, if they haven't logged in within 20 minutes, the agent sends a voice alert prompting them to, and that has significantly cut down our unassigned drive time.' The agent handled more than 1,200 driver-identification calls in a 30-day period.

Catch a Drivers' hours violation three hours before it happens

A Drivers' hours violation usually shows up on a dashboard after the fact, as a compliance record and a fine, not as a chance to prevent it.

Samsara’s AI can analyse a driver's remaining stops, drive time, and shift time, and then flag an approaching violation hours ahead of time so a manager can make a call, like authorising a short-haul extension, before it happens.

(newly approved): 'The agent I built in Samsara Agent Studio tracks remaining stops, drive time, and shift time to flag Drivers' Hours risk three hours out, so we never have to fix non-compliance after the fact,' said Paul Theisen, Safety Coordinator at S. Bertram, a short-haul trucking fleet.

Built for the whole operation

Whether it’s in safety, maintenance, compliance, dispatch, frontline support, or another part of your Organisation, every workflow starts with the same foundation: live operational data, enough context to know what matters, a clear next step, and the right controls.

Because Samsara already captures the day-to-day data teams use to run their operation – from vehicle and asset activity to trips, safety events, maintenance, and more – teams can start with a pre-built capability, tailor the experience in Agent Studio, or bring that trusted Samsara context into the AI tools they already use through MCP – all without building an AI stack from scratch.

The fastest path to value is not transforming everything at once. Start with one workflow your team repeats every day: coaching follow-up, recurring reporting, driver questions, maintenance triage, or the morning briefing. Configure it once, prove it works, and expand from there.

One use case can deliver a practical payoff: less follow-up, fewer manual reports, or faster answers. As more use cases scale across the operation, those gains compound – creating more capacity, faster response, and less manual work without adding people at the same rate.