Connecting travelers with new road trip adventures
Reduction in TCO
Minutes to run a data pipeline vs. 2–3 weeks
Reporting vs. weekly
With a mission to sustainably connect millions with personalized local adventures, Tourism Holdings Limited (thl) — New Zealand’s premier tourism company specializing in camper van and motor home rental across the world — looked to modernize their data platform and unlock new analytical capabilities to better serve their customers and meet growing demands as road trips and camper rentals surged.
As they continued to grow their operations across the world, their legacy on-premises SQL Server system buckled, creating data delays, inaccuracy and security issues. Partnering with Fujitsu Data & AI, thl migrated to Azure Databricks via Fujitsu Data & AI’s Lakehouse Solution Accelerator for its unified lakehouse architecture, collaborative interface and lower total cost of ownership.
With lakehouse architecture in place, thl is able to tap into all of their operational and vehicle telematics data, opening new roads for analytics and machine learning (ML) use cases — from optimizing fleet management to improving daily operations — that will help them steer toward data-driven success and continue connecting millions with unforgettable travel experiences.
A roadblock to operationally efficient growth
Since the COVID-19 pandemic, the love for local exploration and adventure has been rekindled, with road trips surging in popularity around the world. Camper travel in particular has emerged as a favored mode of embarking on journeys that blend freedom, flexibility and immersive experiences. Amid this travel renaissance, thl stands as a trailblazer, orchestrating the transformation of road trip aspirations into tangible realities with their fleet of motor homes, camper vans and caravans. With thousands of vehicles around the world generating data every two seconds per vehicle, thl’s rapid expansion posed major data challenges for their legacy, on-premises platform built on SQL Server. “Our ambitious growth strategies were hindered by a range of data-related obstacles that called for innovative solutions,” said Krishna Pathri, Head of Data and Insights at thl.
One significant hurdle created by their rigid and complex infrastructure was the delayed availability of data for business reporting, resulting in incident calls reported to the data team. The acquisition of various companies over time compounded the issue, leading to a fragmented data infrastructure characterized by data silos and a lack of cohesive strategy. The integration of additional regions was a time-consuming endeavor, further exacerbated by maintenance, security and bug-related challenges. The complexity of semi-structured data worsened the situation, making it difficult to manage within their former infrastructure. “Our previous architecture struggled to scale, causing complications in data management and security,” explained Pathri. “Furthermore, its inability to support ML blocked the potential for us to extract the most value from our data.”
Slow time to value, data accuracy and adequacy emerged as an additional set of problems that impacted decision-making. Departments like finance and marketing weren’t able to make confident decisions, which was evident from documented problems and anecdotes shared within the organization. The complexity of data integration forced the addition of layers of logic, resulting in a convoluted system that affected data accuracy. Slow data onboarding further jeopardized accuracy, posing a formidable challenge for thl. “Our legacy platform’s limitations held us back from expanding our data capabilities,” said Pathri. “We needed to simplify our approach and modernize our platform to support current and future business use cases.”
Navigating more efficient data and analytics with the the Databricks Data Intelligence Platform
To facilitate their strategic vision to be more data-driven across the business, thl partnered with Fujitsu Data & AI to evaluate leading data platforms and tooling, including a multicloud data warehouse and AWS native services. While these alternatives presented certain advantages, thl found them falling short of the holistic solution they required. Pathri recalled, “While each option had its merits, Databricks stood out as the most promising candidate to address our multifaceted data challenges. Ultimately, we wanted a single platform that was future-proof and unlocked new innovations with AI.” Making the choice easier, Fujitsu Data & AI’s Lakehouse Solution Accelerator provides a packaged offering that helps rapidly deliver value from data without worrying about the complexities of implementation.
Leveraging the comprehensive capabilities of the Databricks Data Intelligence Platform, thl has witnessed a transformative impact on their data operations. Databricks SQL empowers analysts to easily explore and query data with their language of choice. Databricks Workflows has streamlined job orchestration, affording greater control over job dependencies and notifications for more efficient data processing. Delta Lake’s ACID transactions and version control have fortified data integrity, providing the ability to revert to previous data states when necessary. The integration of Power BI has elevated reporting and visualization capabilities, currently driving over 15 operational dashboards and reports, with a vision to expand to more in the near future. And with Unity Catalog, Pathri’s team is able to implement fine-grained access controls, an essential requirement for thl’s multi-brand, multi-region setup.
Driving toward operational success and AI innovation
In the ever-evolving landscape of road trip adventures, thl has leveraged the Databricks Data Intelligence Platform and Fujitsu Data & AI’s expertise to transcend data limitations. “Fujitsu Data & AI’s collaboration with Databricks has helped thl to harness the full potential of their operational and telematics data, paving the way for transformative data-driven use cases and unlocking exciting new opportunities through AI,” said Shane Kavanagh, Associate Director at Fujitsu Data & AI.
From a data management standpoint, the Databricks Data Intelligence Platform has served as a catalyst for optimized data processing, reducing time to insight for the business. Prior to using Databricks, thl’s data pipelines could take up to 2 to 3 weeks to run. Now, the same pipelines can run in 45 minutes. With data flowing downstream faster, the business is able to generate reports in a shorter amount of time — from daily to hourly.
Looking ahead, Pathri is excited about the prospect of deploying machine learning use cases to the market that will capture new customers and revenue. “Now that we have a unified platform at our disposal, we are looking at new ML use cases including optimizing fleet utilization by connecting the right vehicles to the right customers, predicting when to repair vehicles to maximize availability and dynamic pricing that responds to real-time signals.” With the Databricks Data Intelligence Platform as their foundation, the road toward data-driven innovation is wide open.
