Posti Group is a modern logistics company with almost 400 years of history. Today, Posti has grown into a company with over €1.4 billion in turnover and more than 13,000 employees, operating across Finland, Sweden, Norway and the Baltics. Through this journey, it plays a vital role in supporting postal services, e-commerce, fulfillment, and last-mile delivery at a national scale, reaching millions of people each day. As customer expectations shift toward faster, more flexible delivery, Posti relies on data and AI to optimize operations, improve service reliability, and innovate across its end-to-end logistics network.
Posti’s journey shows what happens when a large, operationally complex organization treats data architecture as a product: fewer systems, clearer ownership, lower costs, and a foundation built for analytics, operations and AI at scale.
Simplifying Architecture with One Governed Platform (and Everything as Code)
Before Databricks Lakehouse, Posti’s data estate had grown into what the team described as a maze. Analytics ran across multiple warehouses, including Oracle, Azure Synapse and custom-built systems, with fragmented ownership and no clear source of truth. Tracing where data came from, or even handling basic troubleshooting, became a complex task.
Posti modernized onto a single lakehouse architecture with Databricks Lakehouse as the analytics engine and Unity Catalog as the governance layer. One platform now runs all analytics, reporting, downstream applications and AI use cases, with clear ownership and a self-service layer for citizen analysts.
Posti also standardized on an everything-as-code approach, treating even permissions as versioned, auditable code. Arnob Khan, Chapter Lead, Data Engineering and Platform, summarized the shift: “We now have everything as code, infra as code, pipeline as code, SQL as code, even grants as code. Nothing is done in the UI without traceability.”
This approach eliminated configuration drift, improved governance and reduced day-to-day operational friction for both engineers and analysts, producing a platform that scales without adding complexity. As a result, Posti saw significant efficiency gains in onboarding new data products.
“Simplifying the data architecture was a strategic priority, not a technical exercise. Over the years, data capabilities had grown in silos across the organization, resulting in a complex and costly landscape that was difficult to govern and slow to change. By consolidating on one governed platform and adopting a new way of working, we created a modern, scalable and future-ready foundation for data and AI, one that enables faster delivery, clearer ownership and sustainable efficiency at enterprise scale,” said Jyri Ehtamo, Director, Data and Analytics.
Preparing for the Future with Databricks Lakehouse and Lakebase Connectivity
Posti’s data team recognizes an industry-wide trend where the boundary between analytics and operational data usage is becoming less rigid, with growing interest in using governed analytical data more directly to support decision-making, operational visibility and selected operational use cases.
Posti approached this deliberately and collaboratively. Core operational systems remained purpose-built and owned by their respective technology and business teams. Databricks Lakehouse today primarily serves as an analytics and decision-support layer, powering more than 600 production-grade dashboards, supporting insights for over 10 operational systems and enabling a small number of internal and customer-facing APIs where a data-driven view is appropriate.
Looking ahead, capabilities such as Lakebase connectivity are emerging as a way to complement existing architectures rather than replace them, supporting a limited set of fit-for-purpose operational data use cases directly on governed lakehouse data, where it makes sense from a performance, governance and ownership perspective without unnecessary duplication.
“We’re seeing that some use cases benefit from being closer to the data platform, but that doesn’t change the role of operational systems. The focus is on understanding where a data-driven approach adds value and enabling those cases while keeping governance and ownership clear,” Arnob explained.
As Databricks matured from a transformation-focused tool into a broader data platform, Posti reassessed how it should be used at enterprise scale. Earlier implementations were largely centered on data management and isolated analytics, with separate metastores and limited governance capabilities.
The newer lakehouse implementation, governed by Unity Catalog, was designed to provide a stable, enterprise-grade analytics foundation that can also support selected operational and AI-adjacent use cases. Rather than fragmenting responsibilities or architectures, this approach emphasizes alignment: analytics, operational insights and future AI use cases working from a shared, well-governed source of data.
“Databricks has been there since the beginning of our data journey, but earlier it was more of a wrangling tool. Now it’s a complete platform,” Arnob said.
“By focusing first on a strong data foundation, with clear governance and ownership, we created the conditions to evolve deliberately beyond analytics. Today, we see that this approach allows us to support a growing range of business and operational use cases, including those adjacent to AI, and realize value at a completely new scale. The key is balance: extending the platform’s impact where it makes sense, while keeping responsibilities, governance and operational stability fully intact,” Jyri continued.
Transforming GIS and Operational Planning with Lakebase and Unified Geospatial Data
Posti’s GIS and fleet management teams worked across disconnected systems, with planning data, location intelligence and operational workflows spread apart. Postal, e-commerce and freight businesses each operated separate planning environments, while some critical datasets still relied on manual Excel-based workflows. As parcel demand and operational complexity grew across Finland, the lack of a unified operational view made collaboration, optimization and development increasingly difficult.
To address this, Posti consolidated geospatial and operational planning data onto the Databricks Platform and introduced Lakebase as a governed Postgres-compatible layer directly inside it. Planners, analysts and applications now work from one trusted source, with no external database infrastructure or custom integration layers. Vector location data now flows directly from Delta tables into GIS and operational planning applications through Lakebase’s Postgres endpoint, dramatically reducing integration overhead while preserving compatibility with existing GIS tools and workflows.
“There are zero integrations between different systems and almost no lag in spinning up the server,” said Mitchell Cromb, Senior Product Owner for GIS and Fleet Management. “You can’t compare Lakebase to the old infrastructure, where you would boot virtual machines, connect through multiple gateways and depend on tooling from the early 2000s just to access operational data.”
The new architecture also enabled Posti to build operational applications directly on the same governed foundation. Maximus, Posti’s first Databricks App, provides customer service representatives with a map-based view of a mail item’s journey through the sorting network. Processes that previously required specialist SQL expertise and manual joins across six or seven datasets can now be completed by customer service staff in minutes. The new capabilities have already delivered transformational improvements in operational performance and customer experience, enabling faster issue resolution, significantly more accurate customer service and more accurate end-to-end visibility across Posti’s operational network.
“Your optimization potential becomes significantly larger when you can analyze the network as a whole instead of isolated business units,” Mitchell explained. “For the first time, we can realistically model national-scale operational scenarios using governed spatial and operational data from one platform.”
Driving Down TCO While Delivering More
By replacing multiple warehouses and lakehouse implementations with Databricks Lakehouse, Posti reduced its monthly data landscape costs while moving faster. As Arnob put it, “we reduced the cost more than 30%, and at the same time we are delivering more solutions faster than before.”
Posti actively tracks total cost of ownership using Databricks system tables and custom dashboards that monitor usage, forecast spend and alert teams in real time. This gives engineers direct visibility into the cost impact of their design choices, both before and after deployment, making cost a first-class engineering signal rather than a lagging financial metric.
Cost control is maintained without limiting growth, and Posti is now targeting a further cut in platform management overhead.
To summarize Posti’s journey, Jyri said, “We didn’t just consolidate or migrate platforms. We deliberately rethought how our data and AI capabilities should be built for the long term. The goal was to create a scalable, future-ready foundation that allows us to deliver value faster at scale and at a much lower cost. Today, that foundation puts us in a strong position to support business growth, exceptional customer experience and operational efficiency as data and AI become even more central to how we operate.”
FAQ: Posti on Databricks
Posti’s data estate had grown across Oracle, Azure Synapse and custom-built warehouses, creating fragmented ownership, multiple points of failure and no clear source of truth. The complexity made data tracing, troubleshooting and delivery of new data products slower and more expensive.