The Challenge

With data siloed across multiple different systems, a claims management system running on PostgreSQL and multiple separate SaaS API data sources such as commercial billing on Xero, payments via Stripe, multi-currency exchange rates and marketing via Facebook, bringing all this data together without Kleene would have meant a multi-month engineering project and a lot of ongoing management.

With the Kleene platform in place, Laka were able to set up automated data pipelines to power both daily reporting and self-serve around products, policies and claims.

The Solution

In Ben’s experience, people who are not data specialists will often gravitate towards the latest buzzwords in technology, which lately lean in the direction of high-availability, low-structure types of data resources (e.g. data mesh, datamart, data lake etc.). Ben feels that for organisations of a similar size to Laka, or even an order of magnitude larger, a consolidated, structured approach to data will net the best results – which is very much the data warehouse paradigm. He notes that the fact that it is well modelled makes it much easier for people who are less data-literate to interact with and that this is especially true around a single source of the truth. Without one, a simple metric can have an entirely different meaning across departments which are both technically correct, but do not align.

The Results

Ingesting data quickly and easily

Ben finds that one of the trickiest bits when handling data is pulling that data together from different sources. Kleene handles the connection, extraction and unpacking of data seamlessly, when it would be tedious and complicated to code manually. Laka use Xero for all of the invoicing for their commercial clients. Ben feels that while coding extracts for Xero to pull data in JSON format for each different use case is perfectly tractable, that it is not an efficient use of his time. It would probably take a month of full-time software development to build an initial connector which is time and cost prohibitive for them, and they end up having to find other ways to get the information they require.

Managing and maintaining connectors

Ben also notes that in previous work environments he has found that ongoing maintenance of connectors to data sources can be problematic. When version changes to the data source occur, they are often not coordinated with the consumers of that data source and that if you do not immediately make corresponding changes – that could add up to a lot of unplanned development effort – that your source of data stops outright.

Transforming data

When looking at transforming the data, Ben finds that being able to do it all in one place is helpful. Without Kleene, he would be performing a similar job, but a lot more manually and across several tools. Whiteboarding the transforms, writing the SQL and then importing that work into a DAG (Directed Acyclic Graph – used for managing dependencies) tool to verify it is all as needed.

Self-service reporting – saving time

One of the main goals for Laka’s data warehouse project was to reduce the volume of inbound report requests into the data team by enabling self-service reporting for their users. It has been running for just under a quarter now and early signs point to fantastic adoption by the wider business, freeing up time for the data team to focus on the more complex queries that will drive greater business value.

Operating at a higher level

Ben finds that there is a tremendous amount of utility in everyone being able to access reliable measures of business metrics and their change over time. These are some fairly simple calculations but were difficult to get an answer to prior to the project with Kleene. With this data available to users throughout the organisation, decisions can be made without having to wait for the data team to compile new reports. Having the Kleene platform in place saves time and allows the business to function at a higher level.

Running a lean data team

With the Kleene platform, all of the manual work around data extraction and building a data warehouse can be fully automated, freeing up time to drive a data strategy that can be transformative for the business.

Read Our Success Stories!