Analytics Tools

Overview

The Analytics tool category gives you read-only access to four types of reports: agent performance metrics, AI chatbot handling analytics, visitor/session statistics by country, and AI token usage tracking. All tools accept date ranges so you can analyze specific periods or compare across time.

get_agent_performance

Returns per-agent performance metrics for a date range, including conversation counts, response times, and customer ratings.

Parameters

How It Works

• Aggregates performance metrics per agent for the given date range.
• Returns data for every agent who handled at least one conversation in the period.
• Metrics include: total conversations handled, average response time, customer satisfaction ratings, and more.
• Useful for evaluating team performance, identifying agents who need support, and recognizing top performers.

Example Prompts

• "How are my agents performing this month?"
• "Show me agent performance for the last 30 days."
• "Who handled the most conversations this week?"
• "What are the average response times for my team?"
• "Give me a performance report for Q1 2026."
• "Which agent has the best customer ratings?"

get_chatbot_analytics

Returns AI chatbot handling analytics for a specific chatbot over a date range, showing how many conversations the AI handled versus the total.

Parameters

How It Works

• Verifies the chatbot belongs to your organization.
• Returns a daily breakdown of total conversation segments and AI-handled segments.
• Also provides aggregated totals for the full period.
• The AI handling rate (AI segments / total segments) tells you what percentage of conversations the chatbot resolved without human intervention.

Example Prompts

• "How is my Support Bot performing this month?" (lists chatbots first to find the ID)
• "What's my chatbot's AI handling rate for the last 30 days?"
• "Show me daily chatbot analytics for March."
• "How many conversations did the AI resolve on its own this week?"
• "Is my chatbot handling more conversations than last month?" (would need two separate calls for comparison)

get_visitor_stats

Returns visitor and session statistics broken down by country for a date range.

Parameters

How It Works

• Returns per-country session counts.
• Also provides the total number of unique countries and total sessions.
• Useful for understanding your audience's geographic distribution, identifying key markets, and spotting trends in traffic.

Example Prompts

• "Where are my visitors coming from this month?"
• "Show me visitor stats by country for the last week."
• "How many total sessions did I have this month?"
• "Which countries have the most visitors?"
• "Give me a geographic breakdown of my traffic for Q1."

get_ai_usage

Returns daily AI token usage for one or more date periods. Useful for monitoring costs and comparing usage across time.

Parameters

How It Works

• Returns daily token usage for each requested period.
• Supports comparing multiple periods side-by-side (e.g., this month vs. last month).
• Token usage includes all AI operations: chatbot conversations, Copilot usage, content generation, etc.
• Useful for budget tracking, identifying usage spikes, and forecasting costs.

Period Examples

Example Prompts

• "How many AI tokens did we use this month?"
• "Compare our AI usage between February and March."
• "Show me daily AI token usage for the last week."
• "Are we using more AI tokens than last month?"
• "Give me a week-by-week breakdown of AI usage for the past 4 weeks."
• "What's our AI cost trend over the last 3 months?"

Common Workflow Patterns

Monthly Performance Review

Combine multiple analytics tools for a comprehensive overview:

1. Agent performance → get_agent_performance
2. Chatbot effectiveness → get_chatbot_analytics
3. Visitor traffic → get_visitor_stats
4. AI costs → get_ai_usage

Example prompt: "Give me a complete performance report for this month: agent metrics, chatbot handling rate, visitor traffic by country, and AI token usage."

Cost Monitoring

Track AI usage trends to manage your budget:

Example prompt: "Compare AI token usage month over month for the last 3 months. Are we using more or less?"

Chatbot Optimization

Use analytics to identify areas for improvement:

1. Check chatbot handling rate → get_chatbot_analytics
2. Find failed conversations → search_conversations (from Conversation tools)
3. Diagnose issues → diagnose_ai_response (from Chatbot tools)
4. Fix knowledge gaps → manage_chatbot_document (from Chatbot tools)

Example prompt: "My chatbot's handling rate seems low. Show me the analytics, find some recent conversations where the AI failed, and help me fix the knowledge gaps."