Product descriptions:
Automotive group Cycle & Carriage (a member of the Jardine Cycle & Carriage Group) is helping shape the future of mobility with a focus on electric vehicle (EV) innovation and exceptional customer experiences. However, legacy systems and fragmented data silos made it difficult for the company to democratize data access, collaborate effectively and scale AI initiatives. With critical insights trapped across disconnected platforms, Cycle & Carriage lacked the agility needed to meet the demands of a rapidly evolving EV market. By partnering with Databricks and implementing Mosaic AI, the company built a modern, scalable foundation for data-driven innovation, deploying AI chatbots, predictive analytics models and real-time decision intelligence across business units. Cycle & Carriage can now deliver faster, more personalized customer journeys, empower employees with real-time insights and continuously innovate at the pace of change.
Legacy systems stalled digital transformation
As a leading automotive group in Southeast Asia, Cycle & Carriage is driven by a passion to create exceptional, people-focused journeys for customers, employees, partners and communities. With a growing emphasis on the future of mobility, including the rapid adoption of EVs, Cycle & Carriage is investing in digital innovation to meet evolving customer expectations and industry trends. According to Rachel Lim, Chief Digital Officer at Cycle & Carriage, “EV adoption is accelerating due to government incentives, growing environmental awareness and improving infrastructure. We’re taking steps to address this.”
To deliver seamless, personalized experiences across their portfolio of more than 10 automotive brands, including world-class manufacturers like Mercedes-Benz and Kia, Cycle & Carriage set out to become a truly data-driven organization. The company sought to enhance omnichannel EV ownership journeys, optimize operational processes and improve customer engagement and lead conversion using the power of analytics and AI.
Today, the automotive group leverages the Databricks Data Intelligence Platform to power several transformative business use cases: (1) AI chatbot solutions to support dynamic customer engagement, (2) predictive analytics models to optimize internal operations and respond more quickly to market changes and (3) real-time insights that drive smarter, faster decision-making across business units.
However, achieving these goals first required overcoming major technology hurdles. Like many organizations with long-standing IT systems, Cycle & Carriage faced fragmented data silos and sources across legacy platforms. This made it difficult to access and operationalize data at scale. With critical information often trapped in disconnected systems, data democratization was especially challenging. Business units were limited in their ability to collaborate and make timely, informed decisions, reducing the organization’s overall agility.
The company lacked a flexible and robust platform capable of supporting their expanding data and AI ambitions. As Rachel shared, “We needed a platform that could not only scale with our business but also help us build trust in user adoption of AI while ensuring responsible use across the organization.” This set the stage for a partnership with Databricks, providing the modern foundation they needed to transform their data-driven vision into reality.
Powering intelligent customer journeys with Databricks Mosaic AI
To power their digital transformation and drive AI adoption across the organization, Cycle & Carriage turned to Databricks Mosaic AI to accelerate the development and governance of their generative AI (GenAI) initiatives.
Cycle & Carriage developed a retrieval augmented generation (RAG) chatbot that improves productivity and customer engagement by tapping into their proprietary knowledge bases (e.g., technical manuals, customer support transcripts, business process documents) — making it easy for employees to search for information via natural language queries that deliver contextual, real-time answers. After evaluating several models, Rachel chose a 70B parameter variant of Llama for its strong performance, resource efficiency and open source variability.
Rather than rely on a single monolithic model, Rachel deployed an AI agent system, optimizing multiple models built from the same base large language model (LLM) to handle specific tasks more efficiently. This approach allowed her team to deliver tailored, high-quality experiences across different business functions.
Several key components of Mosaic AI were leveraged to build the GenAI solution. Overall, “Mosaic AI provides comprehensive evaluation tools that help assess model quality throughout the development lifecycle,” Rachel said. “This includes using automated LLM-as-a-judge and also storing historical conversations into inference tables for future analysis and tuning,” Rachel explained.
Vector Search was critical to Cycle & Carriage’s RAG architecture, enabling rapid indexing and retrieval of relevant internal documents to enhance the accuracy and contextual relevance of GenAI responses. The Mosaic AI Agent Framework and Agent Evaluation tools allowed the team to orchestrate multiple models and tools into a cohesive, task-driven system. With Model Serving, the company streamlined the deployment of their base and fine-tuned LLMs through low-latency, scalable endpoints, while Mosaic AI Gateway provided centralized control over API access, usage analytics and the flexibility to implement rate-limiting controls as needed.
Mosaic AI Model Training also accelerated their experimentation cycles. “With access to scalable compute and enterprise data stored in the lakehouse, we could iterate on fine-tuning and hyperparameter optimization more efficiently,” Rachel said. While Cycle & Carriage plans to explore additional Databricks features such as Feature Engineering, AutoML, Lakehouse Monitoring and Managed MLflow in the future, the foundation built with Mosaic AI and core platform services has already provided a scalable environment for their AI innovation.
By combining a flexible agentic AI architecture with robust enterprise-grade governance and monitoring capabilities, Cycle & Carriage has created a modern data and AI foundation ready to scale alongside their business ambitions.
Accelerating time to insight for improved customer experiences
Since implementing Databricks Mosaic AI, Cycle & Carriage has seen significant gains in both speed and quality across their AI initiatives. “Databricks Mosaic AI significantly reduced our time to model by streamlining deployment,” Rachel said. “We also observed a notable increase in model accuracy and relevance, improving the user experience in general.”
The RAG system, powered by Databricks’ Vector Search and Agent Framework, has enabled Cycle & Carriage to surface highly contextual, real-time information for employees and customers. This has enhanced productivity across departments by empowering business users with faster, more precise insights without needing deep technical expertise. During this journey, Rachel found great support in Databricks Professional Services. “We worked closely with Databricks solution architects during the development of our GenAI solution. Their support was incredibly valuable as they provided deep technical expertise and guided us through best practices for model training, deployment and governance.”
Additionally, with centralized governance through Mosaic AI Gateway and scalable, low-latency serving infrastructure, Cycle & Carriage has maintained strict quality and compliance standards while still accelerating their experimentation cycles. Internal trust in AI solutions has grown, helping foster wider adoption across the organization.
Today, Cycle & Carriage is better equipped to navigate the fast-changing mobility landscape, delivering more agile customer experiences, smarter operational decisions and an innovation roadmap built to evolve alongside the future of transportation.