# Observe.AI MCP for AI Agents AI Agent Connect > Observe.AI MCP connects your contact center data to your AI client. Pull call transcripts, QA scores, and coaching logs instantly to analyze conversation intelligence without switching tabs. ## Overview - **Category:** customer-support - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_Yt1IMcTvBNoJxvLVhLpLcL7a2RoEMKVbzjuBuyhf/ai-agent-connect - **Tags:** conversation-intelligence, contact-center, quality-assurance, speech-analytics, call-transcription, performance-coaching ## Description Observe.AI gives you a direct line to your contact center performance data through your AI client. Instead of jumping between different dashboards to find a specific call, you can just ask your agent to pull the transcript or summary for you. It pulls in the actual text of chats and emails, along with quality assurance scores and identified business moments like objections or greetings. You can check how your team is doing on specific rubrics or see how often a certain agent is getting coached. This helps you spot trends in customer sentiment or agent behavior much faster than manual auditing ever could. Since Vinkius makes it easy to connect these types of tools, you can jump straight from a high-level question to a specific transcript in seconds. You get a clear view of how your team is actually performing, which helps you make better decisions on training and staffing without the clicking headache. ## Tools ### list_qa_evaluations Get a list of all quality assurance scores across your platform. Use this to find high or low performing interactions quickly. ### get_interaction_details Get the metadata and core info for a specific call or chat. Use this to quickly identify the length and sentiment of an interaction. ### get_interaction_transcript Fetch the full text of a conversation. This is the best way to let your agent read the actual words spoken by the customer. ### list_coaching_sessions See a list of all recorded coaching sessions for your agents. It helps you track how often and with whom your team is being mentored. ### list_evaluation_forms See the different rubrics and forms used for quality checks. This helps you understand the criteria being used to grade your staff. ### list_interactions Get a list of all calls, chats, and emails in your system. This is your starting point for finding any specific customer conversation. ### list_interaction_moments Find specific parts of a call like greetings or objections. This lets you see exactly when a customer becomes frustrated or asks a question. ### list_interaction_summaries See the AI-generated summaries for your recent conversations. It's a fast way to get the gist of dozens of calls in one go. ### list_workspace_users Get a list of all agents and supervisors in your workspace. Use this to identify the right people for specific coaching tasks. ### get_evaluation_details Pull specific data points from a quality assurance evaluation. This helps you see exactly where an agent succeeded or failed a specific rubric. ## Prompt Examples **Prompt:** ``` Find the transcript for the call where the customer was complaining about the login portal. ``` **Response:** ``` I've located the interaction regarding the portal login. Here is the transcript: **Customer:** I'm trying to get into my account but the password reset button isn't doing anything. **Agent:** I'm sorry to hear that. Let me look into that for you. The customer mentioned three times that the 'password reset' button wasn't working. Would you like me to find other calls with this issue? ``` **Prompt:** ``` What was the QA score for the last interaction with agent Sarah? ``` **Response:** ``` Sarah's most recent interaction received a **92% QA score**. * **Strengths**: Empathy and Resolution * **Areas for Improvement**: Compliance (missed a required disclosure) Would you like to see the full evaluation details? ``` **Prompt:** ``` Summarize the main reasons for customer frustration in today's calls. ``` **Response:** ``` Based on the last 20 interaction summaries, here are the main friction points: | Issue | Frequency | | :--- | :--- | | Billing Issues | 40% | | Shipping Delays | 30% | | Account Lockouts | 20% | I can pull more specific transcripts for any of these categories. ``` ## Capabilities ### Pull call transcripts Get the full text of any chat or phone interaction for a deep review of the conversation. ### Check QA scores See specific evaluation metrics and agent performance ratings instantly. ### Find coaching logs Look up previous feedback sessions to see how your agents are improving over time. ### Summarize interactions Get a quick overview of a conversation's main points and outcomes without reading the whole thing. ### Identify key moments Spot specific parts of a call like greetings or objections automatically. ### List workspace users See who is currently active as an agent or supervisor in your instance. ## Use Cases ### QA Audit An analyst needs to find all calls with negative sentiment from yesterday. They ask the agent to list interactions and then pull transcripts for the ones marked as frustrated. ### Coaching Prep A supervisor wants to see if an agent has been coached on billing issues. They ask the agent to list coaching sessions and check the summaries for those specific topics. ### Standup Briefing A manager wants a quick summary of the morning's calls. They ask the agent to list interaction summaries and identify the top three customer complaints. ### Performance Review A coach needs to see an agent's last five QA scores. They ask the agent to list QA evaluations and get the evaluation details for the lowest-scoring one. ## Benefits - Stop manually searching for transcripts by using get_interaction_transcript to pull full conversation logs instantly into your chat. - Spot coaching trends faster by using list_coaching_sessions to see how many times a specific agent has received feedback recently. - Audit quality scores in seconds with list_qa_evaluations instead of clicking through dozens of pages in a web portal. - Get a bird's eye view of customer sentiment by using list_interaction_summaries to see the main themes of your last 50 calls. - Pinpoint specific friction points by using list_interaction_moments to see exactly when customers are raising objections. - Understand the grading criteria clearly by using list_evaluation_forms to see the rubrics your QA team is following. ## How It Works The bottom line is you get a direct line to your conversation intelligence data without leaving your chat interface. 1. Subscribe to this Connector and add your Observe.AI Bearer Token to your configuration. 2. Connect your preferred AI client to the Vinkius catalog to activate the connection. 3. Ask your agent to find transcripts, summarize calls, or pull QA scores using natural language. ## Frequently Asked Questions **Can the Observe.AI MCP help my team find specific call transcripts?** Yes, it lets your AI agent pull the full text of any call, chat, or email instantly. You can ask for a specific interaction ID or search for conversations based on the customer's problem. **How does this Connector help with quality assurance?** It allows you to pull QA scores and evaluation details directly into your chat. You can quickly see how agents are performing against your specific rubrics without opening the Observe.AI dashboard. **Can I use this to see how my agents are being coached?** Yes, you can list all coaching sessions and review the feedback logs. This makes it easy to track an agent's progress and see which coaching topics have been covered. **Does the Observe.AI MCP provide summaries of customer calls?** It does. Your agent can pull AI-generated summaries for recent interactions, giving you a quick way to understand the gist of many calls in seconds. **Is this Connector good for identifying common customer objections?** Definitely. You can use it to list interaction moments, which helps you see exactly when customers are raising specific objections or asking questions during a call. **Can my AI agent see who is in my Observe.AI workspace?** Yes, the Connector can list all agents, supervisors, and admins in your instance. This helps you quickly identify the right people for coaching or reporting. **How do I get an Observe.AI API Key?** You can find your API Key in the Observe.AI dashboard under Settings > API Keys. This token is used as a Bearer Token for authentication. **Can I read full call transcripts through the agent?** Yes! Use the `get_interaction_transcript` tool with a specific Interaction ID to retrieve the full text content of the conversation. **What information is included in the QA evaluations?** Evaluations include the QA form used, individual section scores, overall score, and specific feedback or comments provided by the evaluator.