Vxceed Streamlines Transport Booking with Intelligent, Conversational Application on Amazon Bedrock

Vxceed creates SaaS solutions that help ground transportation providers and global consumer packaged goods (CPG) companies drive sales success, revenue growth, and profitability. Beyond the CPG space, Vxceed also serves industries such as transportation and logistics. One of its flagship offerings, LimoConnect247, is a passenger ground transport application used by government agencies in Australia and New Zealand. The platform provides a unified view of bookings, driver and vehicle availability, real-time location tracking, flight information, scheduling, planning, and alerts.

In 2024, hoping to enhance customer experience and improve operational efficiency, Vxceed launched a strategy to integrate generative AI into its solutions. Cyril Ovely, chief technology officer and co-founder of Vxceed, explains, “We set out to create a conversational application powered by generative AI to better understand human intent. Imagine if LimoConnect247 users could simply state their needs instead of navigating multiple menus. LimoConnect247 holds extensive data—vehicle fleets, drivers, health and safety details, logbooks, and more. Our goal was to make this information easily accessible through natural, intuitive queries.”

Vxceed has long used Amazon Web Services (AWS) as the backbone of its technology stack. For its latest project, the company utilized generative AI on AWS, specifically adopting Amazon Bedrock to develop foundation models (FMs) powered by the Anthropic Claude large language model (LLM). Amazon Bedrock helps users customize FMs with techniques like retrieval-augmented generation and create agents that interact with enterprise data sources and systems. “We decided to use Amazon Bedrock because we’re already heavily invested in AWS, and the service is secure and private,” says Ovely. “Additionally, Amazon Bedrock simplifies collaboration between agents, allowing us to quickly develop a robust, scalable solution.”

Using Amazon Bedrock, Vxceed developed LimoConnectQ, a conversational app where users can create, edit, and cancel ground transport bookings, request quotes, and check existing bookings through text or voice commands. At the core of LimoConnectQ’s multi-agent architecture is an orchestrator—a centralized AWS Lambda function powered by an Amazon Bedrock LLM that interprets user intent.

Accessible via LimoConnect247, LimoConnectQ streamlines the booking process by understanding frequent pickup and drop-off locations, service preferences, and vehicle choices, eliminating the need for lengthy or repetitive chat prompts. The app intelligently identifies whether users need return trips, hourly bookings, or point-to-point trips, looks up flight details for airport pickups, and supports a conversational Q&A feature. This feature helps users access data stored in LimoConnect247, such as policy documents, billing guidelines, contracts, FAQs, and training materials, delivering instant, context-aware responses.

“LimoConnectQ provides a service similar to that of a personal booking agent,” says Ovely. “Instead of navigating menus, users can simply ask, ‘Could you book a trip for Joe from home to work at 10:00 AM tomorrow?’ The app handles the booking, displays trip details, and allows changes if needed.”

The LimoConnectQ service simplifies the booking process, reducing response times for staff inquiries in large transport operations. “Our internal data shows that LimoConnectQ helps staff save at least 10 minutes per booking update and 15 minutes per query—time that would otherwise be spent searching for policies, procedures, or seeking guidance from senior colleagues,” Ovely notes. The app has also improved the first-call resolution rate for staff queries by 80 percent through contextual, data-driven answers. Meanwhile, LimoConnectQ has enhanced staff training by cutting onboarding time by 50 percent. Thanks to its centralized, AI-powered knowledge base, new employees can ask questions and get instant, accurate answers without consulting lengthy manuals or relying on peers.

Security was a high priority in developing LimoConnectQ; many of Vxceed’s government customers are bound by stringent regional data privacy laws, and all data, including booking conversations, needed to be stored in a private, encrypted environment. “Unlike generic public generative AI services, LimoConnectQ doesn’t save, move, or use your data,” says Ovely. Vxceed is enhancing LimoConnectQ by using Amazon Bedrock Guardrails and adding features such as a conversational interface tailored for dispatchers and travel schedulers. “We’re just getting started on our generative AI journey, and we’re excited to continue working with AWS to meet our goals,” Ovely concludes.