Skip to main content
HoneyHive Chart Interface Example

​
Quick Start: Building Your First Chart

Creating insightful visualizations in HoneyHive is straightforward.
1

Access Discover

Click New Chart in your Dashboard, or navigate to the Discover tab from the sidebar.
2

Select Your Data Source

Choose from three data scopes:
ScopeWhat it covers
SessionsFull user interactions/traces (entire conversations)
CompletionsIndividual LLM calls
All EventsAny tracked step in your pipeline, including tool calls
3

Configure Your Visualization

SettingDescription
EventWhich event type to analyze (default: All Sessions/Completions/Events)
MetricWhat to measure (e.g., Request Volume, Duration, Cost, or custom evaluators)
AggregationHow to calculate (Sum, Average, Median, 99th Percentile, etc.)
4

Refine Your Analysis (Optional)

SettingDescription
FilterNarrow down to specific data segments (e.g., source = "production")
Group BySplit results by properties (e.g., prompt_version, model, user_tier)
Time RangeSet your analysis window (1d, 7d, 30d, etc.)

​
Understanding Your Data

To build effective charts, it helps to understand the data components available in HoneyHive.

​
Metrics

Metrics are the numerical values you visualize in charts.
MetricWhat it tells you
Request VolumeQueries over time. Spot usage spikes or drops.
CostDirect expenses. See if that new feature is breaking the bank.
DurationSystem latency. Slow responses kill engagement.
Your custom quality checks, either Python or LLM-based. Must return float or boolean to chart.
ExampleTypeQuestion it answers
Keyword PresencebooleanDoes every product review mention the product?
Coherence ScorefloatHow logically sound are multi-turn conversations?
The voice of your users, quantified. Accepts float or boolean inputs.
ExampleTypeQuestion it answers
Usefulness RatingfloatOn a scale of 1-5, how useful was this response?
Used in ReportbooleanDid the user actually use this in their report?

​
Properties

Properties provide context for your metrics. All properties in the enrichment schema such as config, user properties, feedback, metrics, and metadata can be used to slice and dice your data.
Metrics chart performance. Properties unveil the context behind that performance. Both are crucial for exploratory data analysis.

​
Chart Types

Each chart type focuses on a different part of your LLM pipeline.
Focus: Individual LLM calls.Key Metrics: cost, duration, tokens, errors, and any specified evaluators.

Example use case

Hypothesis: “Longer user messages cause more token waste.”Test: Chart Average Unused Output Tokens grouped by binned_input_length.

​
Next Steps

Set Up Alerts

Get notified when metrics exceed thresholds

Online Evaluations

Run evaluators automatically on production traces