Analytics API

Access usage analytics, performance metrics, and query insights for your bots.

Get Bot Analytics

GET/analytics/bots/{agentId}/

Get comprehensive analytics for a specific bot including query counts, response times, and popular questions.

Query Parameters

periodstringdefault: week

Time period: day, week, or month

Example Request

-purple-500">curl https://api.ragchats.ai/api/analytics/bots/AGENT_ID/?period=week \
  -H -green-500">"Authorization: Bearer YOUR_TOKEN"

Response

{
  "totalQueries": 1250,
  "averageLatencyMs": 890,
  "successRate": 0.98,
  "topQueries": [
    {
      "query": "What is the return policy?",
      "count": 45
    },
    {
      "query": "How do I track my order?",
      "count": 38
    },
    {
      "query": "What payment methods do you accept?",
      "count": 32
    },
    {
      "query": "How do I contact support?",
      "count": 28
    },
    {
      "query": "What are your shipping options?",
      "count": 25
    }
  ],
  "queriesPerDay": [
    { "date": "2024-01-08", "count": 150 },
    { "date": "2024-01-09", "count": 175 },
    { "date": "2024-01-10", "count": 200 },
    { "date": "2024-01-11", "count": 180 },
    { "date": "2024-01-12", "count": 165 },
    { "date": "2024-01-13", "count": 190 },
    { "date": "2024-01-14", "count": 190 }
  ]
}

Response Fields

BotAnalytics

totalQueriesinteger

Total number of queries in the period

averageLatencyMsnumber

Average response time in milliseconds

successRatenumber

Percentage of successful responses (0-1)

topQueriesarray

Most frequently asked questions

querystring

Question text

countinteger

Number of times asked

queriesPerDayarray

Daily query counts

datedate

Date (YYYY-MM-DD)

countinteger

Queries on that date

Understanding Metrics

Total Queries

The total number of queries sent to your bot during the selected period. This includes both successful and failed queries.

Average Latency

The average time in milliseconds from when a query is received to when the response is fully generated. This includes:

Document retrieval time
Reranking (if enabled)
LLM response generation

Target: Under 2,000ms for good user experience.

Success Rate

The percentage of queries that returned a successful response. A query is considered successful if:

No errors occurred
Relevant context was found
An answer was generated

Target: Above 95%.

Top Queries

The most frequently asked questions. Use this to:

Identify common customer needs
Ensure documentation covers these topics
Improve answers for popular questions

Period Options

dayperiod

Last 24 hours, hourly breakdown

weekperiod

Last 7 days, daily breakdown

monthperiod

Last 30 days, daily breakdown

Use Cases

Build a Dashboard

500">class=500">class="text-green-500">"text-muted-foreground">// Fetch analytics 500">for dashboard
500">async 500">function getAnalytics(agentId, period = 500">class="text-green-500">'week') {
  500">const response = 500">await fetch(
    500">class="text-green-500">`https:500">class="text-muted-foreground500">class="text-green-500">">//api.ragchats.ai/api/analytics/bots/${agentId}/?period=${period}`,
    {
      headers: { 500">class="text-green-500">"Authorization": 500">class="text-green-500">`Bearer ${token}` }
    }
  );
  500">return response.json();
}

500">class=500">class="text-green-500">"text-muted-foreground">// Display metrics
500">const analytics = 500">await getAnalytics(500">class="text-green-500">'bot_123', 500">class="text-green-500">'week');
console.log(500">class="text-green-500">`Queries 500">this week: ${analytics.totalQueries}`);
console.log(500">class="text-green-500">`Avg latency: ${analytics.averageLatencyMs}ms`);
console.log(500">class="text-green-500">`Success rate: ${(analytics.successRate * 100).toFixed(1)}%`);

Monitor Performance

500">class=500">class="text-green-500">"text-muted-foreground">// Alert 500">if latency is too high
500">const analytics = 500">await getAnalytics(500">class="text-green-500">'bot_123');

500">if (analytics.averageLatencyMs > 3000) {
  sendAlert(500">class="text-green-500">'Bot latency is high: ' + analytics.averageLatencyMs + 500">class="text-green-500">'ms');
}

500">if (analytics.successRate < 0.95) {
  sendAlert(500">class="text-green-500">'Bot success rate dropped to ' + (analytics.successRate * 100) + 500">class="text-green-500">'%');
}

Identify Content Gaps

500">class=500">class="text-green-500">"text-muted-foreground">// Find questions that might need better documentation
500">const analytics = 500">await getAnalytics(500">class="text-green-500">'bot_123', 500">class="text-green-500">'month');

console.log(500">class="text-green-500">'Top questions to review:');
analytics.topQueries.slice(0, 10).forEach((q, i) => {
  console.log(500">class="text-green-500">`${i + 1}. "${q.query}500">class="text-green-500">" (asked ${q.count} times)`);
});