1# Eval Correlation — Find Evaluation Results by Response or Conversation ID
2 
3Look up evaluation scores for a specific agent response using App Insights.
4 
5> **IMPORTANT:** The Foundry evaluation API does NOT support querying by response ID or conversation ID. App Insights `customEvents` is the ONLY way to correlate eval scores to specific responses. Always use this KQL approach when the user asks for eval results for a specific response or conversation.
6 
7## Prerequisites
8 
9- App Insights resource resolved (see [trace.md](../trace.md) Before Starting)
10- A response ID (`gen_ai.response.id`) or conversation ID (`gen_ai.conversation.id`) from a previous trace query
11 
12## Search by Response ID
13 
14```kql
15customEvents
16| where timestamp > ago(30d)
17| where name == "gen_ai.evaluation.result"
18| where customDimensions["gen_ai.response.id"] == "<response_id>"
19| extend
20    evalName = tostring(customDimensions["gen_ai.evaluation.name"]),
21    score = todouble(customDimensions["gen_ai.evaluation.score.value"]),
22    label = tostring(customDimensions["gen_ai.evaluation.score.label"]),
23    explanation = tostring(customDimensions["gen_ai.evaluation.explanation"]),
24    responseId = tostring(customDimensions["gen_ai.response.id"]),
25    conversationId = tostring(customDimensions["gen_ai.conversation.id"])
26| project timestamp, evalName, score, label, explanation, responseId, conversationId
27| order by evalName asc
28```
29 
30## Search by Conversation ID
31 
32```kql
33customEvents
34| where timestamp > ago(30d)
35| where name == "gen_ai.evaluation.result"
36| where customDimensions["gen_ai.conversation.id"] == "<conversation_id>"
37| extend
38    evalName = tostring(customDimensions["gen_ai.evaluation.name"]),
39    score = todouble(customDimensions["gen_ai.evaluation.score.value"]),
40    label = tostring(customDimensions["gen_ai.evaluation.score.label"]),
41    explanation = tostring(customDimensions["gen_ai.evaluation.explanation"]),
42    responseId = tostring(customDimensions["gen_ai.response.id"])
43| project timestamp, evalName, score, label, explanation, responseId
44| order by responseId asc, evalName asc
45```
46 
47## Present Results
48 
49Show eval scores as a table:
50 
51| Evaluator | Score | Label | Explanation |
52|-----------|-------|-------|-------------|
53| coherence | 5.0 | pass | Response is well-structured... |
54| fluency | 4.0 | pass | Natural language flow... |
55| relevance | 2.0 | fail | Response doesn't address... |
56 
57When showing alongside a span tree (see [Conversation Detail](conversation-detail.md)), attach eval scores to the span whose `gen_ai.response.id` matches.
58