Today we're launching Fleetbase AI: the open-source AI for logistics, built directly into the operations platform and live now. Ask about your operation or describe an action in plain English, and Fleetbase AI takes it from there — inside your workflow, on infrastructure you control.
Most "AI-powered" logistics tools are black boxes bolted onto a closed product. Fleetbase AI is a different kind of logistics AI: open, auditable, model-agnostic, and embedded in the platform you already run. As far as we know, it's the first open-source logistics AI shipped like this.
Logistics AI is broken - and you're paying for it
If you've evaluated the "AI" features in most logistics platforms, you know the pattern. You get a recommendation with no explanation. Your operational data is shipped off to a third-party AI provider. The insights live in some separate dashboard, away from where the work actually happens. And the moment you depend on their AI, you're locked into their pricing, their uptime, and their roadmap.
That's not intelligence you own — it's a liability you rent.
A different kind of logistics AI
Fleetbase AI is open, auditable, and embedded directly in Fleet-Ops and the wider platform. It acts inside your workflow, not in a silo beside it. Four things make it different:
- Bring your own model. OpenAI, Claude, or a local model — Fleetbase AI is model-agnostic. No forced dependency on a single provider's pricing or roadmap.
- Every action previews before it applies. Nothing happens silently. You see the draft, you approve it, then it goes live.
- Fully audited. Every AI task, step, and session is recorded. Admins gate which providers and configuration are allowed, and disabled AI is blocked server-side.
- Open source, on infrastructure you control. Read the module, see exactly how it reasons over your operations, self-host it, and build on it.
How it works
AI sees your operation. Modules register AI-readable context — orders, drivers, vehicles, routes, and help docs — so Fleetbase AI works from your live operational data, not a generic model guessing at your business.
You ask in plain English. Open the AI prompt from the console header (or a keyboard shortcut away) and ask a question or describe an action. Multi-turn chat keeps the context, so follow-ups just work.
AI acts in the platform. Fleetbase AI doesn't just reply — it can draft orders, run queries, and trigger actions directly inside Fleetbase. Every action previews first.
You stay in control. Every task, step, and session is recorded for audit. Admins decide what's allowed; disabled AI is fully blocked server-side.
See it in action
Here's Fleetbase AI turning a single plain-English instruction into a ready-to-dispatch order draft — addresses resolved, pins on the map, order type set — all live in Fleet-Ops.
What it means for the people running the fleet
For a dispatch manager, the win is speed and consistency. Type "create an order from Marina Bay Sands to 42 Hougang Avenue 7 and assign whoever's online" and Fleetbase AI resolves the addresses, plots the route, sets the order type, and hands you a ready-to-confirm draft. Orders get created the same way every time — the handoff to a driver takes seconds instead of minutes.
For an operations manager, it lowers the floor on training and cuts errors. A new coordinator doesn't need to memorize where every field lives — they describe the outcome and review a structured draft. Fewer mis-keyed addresses, fewer orphaned orders, fewer "who was supposed to take this?" moments.
For a fleet manager, it's leverage with accountability. The same headcount handles more volume, and because every AI-prepared action is a preview you approve — and every step is audited — you keep a clear line of sight over what actually happens in the field.
What Fleetbase AI does today
Live capabilities, built into the platform and shipping in the open:
- A global AI prompt in your console header, or a keyboard shortcut away — ask about your operation, get answers, and take action in plain English.
- Natural-language order creation — describe a job and get a previewed order draft.
- Order insights and operational queries — ask questions across your live data.
- AI route optimization.
- An AI framework every extension can plug into, so capabilities grow as the platform does.
Open source is a competitive advantage
Closed-source logistics AI is a liability. Open-source AI is an asset you can build on, audit, and own. With a black box, you have no idea why a route was chosen or an action was taken. Fleetbase AI is open — read the module, see the reasoning, run it on your own infrastructure, and extend it to fit how your business actually works.
- Fleetbase AI module: https://github.com/fleetbase/ai
- Release v0.7.49: https://github.com/fleetbase/fleetbase/releases/tag/v0.7.49
- Release v0.7.51: https://github.com/fleetbase/fleetbase/releases/tag/v0.7.51
Try it
Fleetbase AI is live on Fleetbase Cloud and self-hosted — enable it in your console. Shipped in v0.7.49. Update to the latest release, open Fleet-Ops, and open the AI prompt from the console header. Describe what you want, and review the draft it prepares.
We're just getting started — and we're building it in the open. Tell us what you'd want to run from a single prompt.
