LLM Application Monitoring & Protection

Monitoring LLM Applications with Fiddler

Monitoring of Large Language Model (LLM) applications with Fiddler requires publication of the LLM application inputs and outputs, which include prompts, prompt context, response, and the source documents retrieved (for RAG-based applications). Fiddler will then generate enrichments, which are LLM trust and safety metrics, for alerting, analysis, or debugging purposes.

Fiddler is a pioneer in the AI Trust domain and, as such, offers the most extensive set of AI safety and trust metrics available today.

Selecting Enrichments to Enhance Monitoring

Fiddler offers many enrichments that each measure different aspects of an LLM application. For detailed information about which enrichment to select for any specific use case, visit this page. Some enrichments use Fast Trust Models to generate these scores.

Generating Enrichments with Fiddler

LLM application owners must specify the enrichments to be generated by Fiddler during model onboarding. The enrichment pipeline then generates enrichments for the LLM application's inputs and outputs as events are published to Fiddler.

Fiddler Enrichment Framework diagram displaying sample inputs and outputs flowing into the Fiddler enrichment pipeline.

Figure 1. The Fiddler Enrichment Framework

After the LLM application's raw unstructured inputs and outputs are published to Fiddler, the enrichment framework augments them with various AI trust and safety metrics. These metrics can monitor the application's overall health and alert users to any performance degradation.

Fiddler dashboard showing LLM application performance using enrichment metrics.

Figure 2. A Fiddler dashboard showing LLM application performance

With the metrics produced by the enrichment framework, users can monitor LLM application performance over time and conduct root cause analysis when problematic trends are identified.

At the time of model onboarding, application owners can opt in to the various and ever-expanding Fiddler enrichments by specifying fdl.Enrichment as custom features in the Fiddler ModelSpec object.

# Automatically generating an embedding for a column named “question”

fiddler_custom_features = [
        fdl.Enrichment(
            name='question_embedding',
            enrichment='embedding',
            columns=['question'],
        ),
        fdl.TextEmbedding(
            name='question_cf',
            source_column='question',
            column='question_embedding',
        ),
    ]

model_spec = fdl.ModelSpec(
    inputs=['question'],
    custom_features=fiddler_custom_features,
)

The code snippet above illustrates how the ModelSpec object is configured to opt in to an embedding enrichment, which is then used to create a fdl.TextEmbedding input. This input allows for drift detection and embedding visualizations with UMAP.

Enrichments Available

Please reference fdl.Enrichment for a list of available enrichments as of the latest release.

---

❓ Questions? Talk to a product expert or request a demo.

💡 Need help? Contact us at [e-mail].