Step 4: Service Operation
Finally, we offer our customers service operation. Just to be clear, we are not a hosting provider! We only offer operation for services or databases that we have created ourselves. We have customers with different requirements. Our approach to building AI services for our customers also includes the need to operate them.
For the most part, we find ourselves in one of the following situations:
- Our customer has the infrastructure on site, i.e. our service has to be installed and integrated – this is acceptable for us.
- Our customer has an infrastructure, but wants us to host our service ourselves (at least temporarily), but integrate it with his Systems (REST, API, …) – this is no problem and is often faster.
- Our customer has no infrastructure of his own or is missing parts of the know-how and wants us to operate the services completely and permanently for him, which we handle gladly.
We then operate the base platform, all functional modules and the REST and Micro services for integrating our algorithms into existing systems. This includes all necessary databases, starting with simple small MySQL or PostgreSQL systems up to Big Data Clusters with MongoDB, ElasticSearch and Cassandra, which can also go into the petabyte range.
We have our own servers in Germany that are also distributed over several locations. In addition, we are familiar with the cloud infrastructure of Amazon AWS, Microsoft Azure and Google Cloud and use these upon customer request.