• AI & Data Science
April 22, 2024

Combine Technology Platform: The way to quicker and better maintained projects

Combine’s internal team has taken on several types of data science projects during the last years. Seeing so many different projects in a short period of time reveals the similarities in the approaches and tools that an engineer will use. This has led to the decision to create our own technology platform, where all these different tools are generalized and maintained, which makes them available for all the different projects. The idea is that this will lead to a quicker start-up of projects, a better collaboration between engineers and projects that are easier to maintain.

Combine’s internal team has taken on several types of data science projects during the last years. Seeing so many different projects in a short period of time reveals the similarities in the approaches and tools that an engineer will use. This has led to the decision to create our own technology platform, where all these different tools are generalized and maintained, which makes them available for all the different projects. The idea is that this will lead to a quicker start-up of projects, a better collaboration between engineers and projects that are easier to maintain.

Similarities between projects

Even though projects can be related to completely different fields, like the shipping industry, ACHT 2 (read more), a desalination device, Avsalt (read more), or our sewage water systems, the way one needs to process the data so that we can work with it, extract features, give insights and train AI on, is remarkably similar. All projects need a structured way to import the data and preprocess it to a format that is easy to work with. Moreover, all projects will need some data quality checks to see if there is data missing, if the values are consistent and whether the signals that you expect to correlate actually do.  

Another common part between the different projects is how you present the data insights. Most industries are not ready to directly jump to a fully automated AI solution and instead need a user interface to show them the insights and build trust over time. 

What the platform consists of

Our Combine Technology Platform (CTP) consists of distinct parts that cover the similarities between the projects, like building blocks for the data science pipeline to import and preprocess the data, automated quality checks, APIs, and a dashboard. 

Besides these building blocks, the platform also contains automated set-ups for new GitLab repositories, building blocks for CI checks and documentation on a streamlined way of working. All these ensure the quality of the projects. 

CTP is lightweight, based on free open-source software and it is possible to run everything locally. This makes this the CTP different from many other services that usually only run in the cloud and combine several licenses/programs to achieve the same goal.  

How can your project profit from this?

Both new projects and current projects profit from the Combine Technology Platform. New projects can experience a much shorter path to production, since we have tools available that are tailored to our way of working and that directly make the project production ready and scalable. This allows you to start using AI in the company’s daily tasks a lot sooner.  

Projects that have already been collaborating with us will profit from it in two different ways. Firstly, the projects and code base are easier to maintain, since they profit from all updates and development that we will do during other projects. This ensures that the code does not degenerate, and that the quality of the projects remains high. Secondly, it profits from the improved collaboration between engineers that the Combine Technology Platform enables. The platform introduces common ways of working and documentation on all discoveries. This enables knowledge transfer between members of the team and towards new members, which ensures a constant quality of the project.