Run Communication Workflows with AI

When high winds force flight cancellations, Papillon may need to contact hundreds of customers in a very short time window—over 400 reservations within 24 hours. In the past, this would have required call center staff working around the clock, manually coordinating outreach, tracking who hadn't been reached, following up on no answers, and rebooking guests into alternate dates. Instead, Papillon now runs a "cancellation and rebooking" workflow where the AI makes concurrent calls, leaves voicemails, sends follow-up text messages, associates callbacks to the reservation, speaks in the customer's language, and continues outreach until each case is resolved, turning what was once an all-hands emergency into a highly efficient, automated process.

Gray Line receives reservations from booking systems around the world, many of which don't capture the pickup details needed for smooth operations, often leading to early-morning logistical issues. Previously, staff would have spent hours manually calling and messaging guests to track down hotel information. Now, Gray Line runs a "collect hotel information" workflow by uploading a customer file and instructing the AI to contact guests concurrently, gather hotel details, and return a structured summary to the team. Once locations are confirmed, staff direct the AI to send pickup instructions immediately and again before the tour, shifting their role from repetitive outreach to overseeing and directing their AI worker.

During routine service visits, technicians often recommend preventative fixes that customers decline at the time, even though delaying them can lead to bigger and more expensive problems later. In the past, these follow-ups rarely happened because front-desk staff couldn't answer detailed technical questions and service technicians didn't have time to make outbound calls. Now, service managers periodically upload a customer file and run a preventative maintenance workflow, where the AI contacts thousands of customers with personalized information from their last visit—explaining the issue, what happens if it's ignored, and the cost difference between fixing it now versus later—while answering questions, collecting intent, and gathering a preferred date range to return, resulting in higher service revenue and helping customers avoid larger repair bills down the road.