Add ADR for using AI Verify for running technical tests

docs/projects/tad/adrs/0005-ai-verify-technical-tests.md

@@ -0,0 +1,39 @@
+# TAD-0005 Use AI Verify for technical tests
+
+## Context
+
+The AI Verify project is set up in a modular way, and the technical tests are one of the modules. The AI Verify
+team is developing a feature which makes it possible to run the technical tests using an API, by using
+a python file calling a method to run a test and providing the required configuration, like which
+model and dataset to use and some test specific requirements. The result of the test is written to a file in a JSON format.
+
+### Pros
+
+* We have several technical tests we can use of the shelf.
+* Tests are set up in a generic way using interfaces which allows others, like ourselves, to create their own plugins.
+* Loading models, pipelines and data is done through the AI Verify toolkit, which does validation
+ and support checks.
+
+### Cons
+
+* The result of a test is dependent on how AI Verify has implemented it.
+* Adding our own plugins may require adding AI Verify front end blocks we don't use ourselves.
+* We are dependent on the AI Verify ecosystem for supported models and data formats.
+
+## Assumptions
+
+* We can wrap the API and other AI Verify requirements in a docker image.
+* We can run the docker image independently where we only have to provide the model, datasets and other
+ required configuration to run a test.
+
+## Decision
+
+We will use the technical tests of AI Verify for the TAD tool.
+
+## Risks
+
+We are tied to the AI Verify ecosystem.
+
+## Consequences
+
+We have a set of technical tests we can integrate in the TAD tool.