LLM Pulse Agent (AI Chat Assistant)

The LLM Pulse Agent is an AI analyst built into the app. Ask it questions about your data in plain language ("how is my visibility trending?", "which sources cite my competitor but not me?") and it answers with your real numbers, interactive charts, and exportable tables. It replies in whatever language you write in.

With the LLM Pulse Agent, you can:

• Ask about your AI visibility, competitors, citations, and sentiment in plain language
• Get answers as real numbers, interactive charts, and exportable tables
• Have it act for you: create prompts, competitors, tags, and annotations on request
• Skip the dashboards when you just need one specific answer fast

Open it from the Agent link at the top of the sidebar (full-page view) or the floating chat bubble in the bottom-right corner of any page. Available on Growth plans and above.

What it does

• Answers data questions: visibility and mention rate trends, share of voice, citations and top sources, sentiment, per-model comparisons, individual AI responses, and more
• Renders charts and tables: timeseries, share of voice, and top sources come back as interactive charts; every table has Copy, CSV, and Excel buttons above it
• Acts on your project when you ask: it can create prompts (single or bulk), add competitors, create tags and collections, tag prompts, post annotations, launch Recommendations runs, and start GEO Writer tasks. It always confirms before creating anything that consumes quota.
• Fills in your project profile: ask it to set your brand description, industry, business model, target audience, primary products, brand voice, goals, or your brand matching names, and it updates them for you after you confirm. Matching names and primary products are replaced as a whole list, so tell it every value you want kept. Changing your brand name or matching names re-runs mention detection over your history, so past numbers can shift. The project name, website, default country and language, and app store or social links stay in Project Settings.
• Answers product and account questions: how metrics are calculated, what your plan includes, how many prompts you have left, when your next data run is scheduled

How to use it

1. Click Agent in the sidebar or the chat bubble in the corner
2. Ask in your own words; no special syntax needed. Examples:
   • "Show me share of voice by competitor for the last 90 days"
   • "Which Reddit threads are cited when people ask about my category?"
   • "Generate 20 Spanish prompts for car buyers and add them to this project"
   • "Why did my visibility drop last week?"
3. Follow up naturally; the agent keeps the conversation context
4. Download any table it produces with the CSV or Excel buttons above the table

Tips & notes

• Every reply shows a thumbs up / thumbs down control below it. Rating replies helps us improve the agent, and clicking the active thumb again clears your rating
• Message quotas depend on your plan and are shared across your team, resetting on the 1st of each month. Growth conversations are capped at 25 of your messages. Scale and higher plans allow 60 messages per conversation. Start a new conversation when you hit the cap.
• The agent only sees the currently selected project, and conversations are never used to train AI models
• If you give no time period, it defaults to the last 30 days and says so
• For data questions it always pulls live numbers from your project rather than estimating
• Prefer using your own AI assistant? Connect ChatGPT, Claude, or any MCP client to your LLM Pulse data instead; see Connect your AI via MCP