Guardrails AI

Guardrails AI is a framework for validating the inputs and outputs

For this guide we will use a simple example that logs guardrails validation steps as traces to Opik, providing them with the validation result tags.

First, ensure you have both opik and guardrails-ai installed:

$ pip install opik guardrails-ai

We will also need to install the guardrails check for politeness from the Guardrails Hub

$ guardrails hub install hub://guardrails/politeness_check

Logging validation traces

In order to log traces to Opik, you will need to call the track the Guard object with track_guardrails function.

1 from guardrails import Guard, OnFailAction
2 from guardrails.hub import PolitenessCheck
3 
4 from opik.integrations.guardrails import track_guardrails
5 
6 politeness_check = PolitenessCheck(
7     llm_callable="gpt-3.5-turbo", on_fail=OnFailAction.NOOP
8 )
9 
10 guard: Guard = Guard().use_many(politeness_check)
11 guard = track_guardrails(guard, project_name="guardrails-integration-example")
12 
13 guard.validate(
14     "Would you be so kind to pass me a cup of tea?",
15 )
16 guard.validate(
17     "Shut your mouth up and give me the tea.",
18 );

Every validation will now be logged to Opik as a trace

The trace will now be viewable in the Opik platform:

1	from guardrails import Guard, OnFailAction
2	from guardrails.hub import PolitenessCheck
3
4	from opik.integrations.guardrails import track_guardrails
5
6	politeness_check = PolitenessCheck(
7	llm_callable="gpt-3.5-turbo", on_fail=OnFailAction.NOOP
8	)
9
10	guard: Guard = Guard().use_many(politeness_check)
11	guard = track_guardrails(guard, project_name="guardrails-integration-example")
12
13	guard.validate(
14	"Would you be so kind to pass me a cup of tea?",
15	)
16	guard.validate(
17	"Shut your mouth up and give me the tea.",
18	);