Mutuall — Chrome Extension for Logistics Scheduling Consolidation

Scheduling a single freight booking already means juggling a supplier portal, the receiving facility, and any intermediaries. Scheduling five loads typically meant logging into roughly fifteen different sites in a single afternoon. Five separate sets of credentials, five separate UIs, five separate copy-paste cycles between PO and slot picker.

The work itself, picking the right slot at the right facility, was minutes of decision-making per load. Everything around it was hours of browser tax.

When we shadowed logistics teams, the actual scheduling decisions (which slot, which carrier, which facility) were a small fraction of their day. Hours of it went into context-switching between portals to execute those decisions. The leverage point wasn't smarter scheduling; it was collapsing the switching cost to zero.

Replacing the portals was a non-starter. We had to live inside them.

Build a single browser-based companion that uses encrypted on-device credentials to dispatch user-approved browser agents across supplier portals. Schedulers approve a recommendation in one surface; the agents handle the repetitive cross-portal execution under human supervision. The user stays in control of every booking.

Encrypted on-device credential storage and an exportable activity log gave the workflow the trust + audit posture logistics IT teams expected before letting any automation touch their portals.

The extension lived entirely in the user's Chrome and ran three jobs. (1) A PO lookup engine matched POs across portals to their correct facility IDs even when the portal UIs didn't surface that mapping. (2) Browser agents executed the cross-portal scheduling work using encrypted on-device credentials. Every action was queued for explicit user approval before execution — agents recommended and acted on direct human go-ahead, never autonomously. (3) An exportable activity log captured every action with audit-grade detail.

Hard architectural constraint: human-in-the-loop on every booking. The product never autonomously booked an appointment. That made the failure mode 'agent didn't book' instead of 'agent booked the wrong thing,' which mattered a lot for adoption inside logistics IT.

No backend. The entire extension lived inside the Chrome process. Credentials never left the user's machine, which removed an entire class of compliance conversations.

Across active users, scheduling time per load dropped by 87%. The same scheduler now did in a morning what used to fill an entire day.

The activity log turned out to be the quiet second win. Logistics teams had been hand-maintaining appointment trackers in spreadsheets for audit purposes; the extension produced that as a side effect, exportable and accurate by construction.

We ran the product for two years and sunset it in 2025 after an amicable wind-down. The Chrome Web Store listing stayed live through the cycle.