Hello there! Ever found yourself wondering, "What did I just say to that AI?" or "Did the Poly AI understand my query correctly last time?" If so, you're in the right place! Understanding how to access and review your interaction history on the Poly AI platform is crucial for troubleshooting, optimizing performance, and simply keeping track of your conversations.
This comprehensive guide will walk you through the process step-by-step, ensuring you can confidently navigate and utilize the historical data available to you. Let's dive in!
Navigating Your Poly AI Journey: How to Check History
Checking history on Poly AI typically involves accessing their platform's "Conversation Logs" or a similar analytical section. The specifics might vary slightly depending on your role (end-user, administrator, developer) and the specific Poly AI product or integration you are using (e.g., Poly AI for customer service, a specific app, or an API integration). However, the general principles remain consistent.
Step 1: Identify Your Poly AI Access Point
Before we delve into the specifics, the very first thing you need to do is identify how you typically interact with Poly AI. Are you:
Using a Poly AI-powered chatbot on a company's website?
Interacting with a voice AI agent over the phone?
A developer or administrator accessing the Poly AI backend platform?
Using a specific Poly AI application like "PolyBuzz: Chat with AI Friends" or "Poly Clariti App"?
Understanding your access point is crucial, as it dictates the method you'll use to check the history.
Step 2: Accessing the Poly AI Platform (If Applicable)
This step applies primarily to administrators, developers, or users of specific Poly AI applications that have a dedicated web portal or user interface.
Sub-heading 2.1: For Web-Based Poly AI Platforms (Administrators/Developers)
If you are an administrator or developer managing Poly AI agents, you'll likely have access to a dedicated Poly AI platform or an integrated system (like Zendesk, if Poly AI is integrated there).
Log In: Navigate to your Poly AI administrator portal or the platform where your Poly AI solution is managed. You'll need your username and password to log in securely.
Locate "Conversation Logs" or "Analytics": Once logged in, look for a menu item or a dashboard section labeled "Conversation Logs," "Chat History," "Analytics," or similar. These sections are specifically designed to provide insights into AI interactions. In some integrated systems (like Zendesk), this might be found under "AI agents - Advanced" and then "Conversation logs."
Select the AI Agent (if multiple): If your organization uses multiple Poly AI agents, there might be a dropdown or selection option to choose the specific AI agent whose conversation logs you wish to view.
Sub-heading 2.2: For Poly AI Applications (e.g., PolyBuzz, Poly Clariti App)
If you're using a specific Poly AI-powered application, the history might be directly accessible within the app itself.
Open the Application: Launch the Poly AI application on your device (e.g., mobile phone, desktop).
Look for Chat History/Conversation Section: Within the app's interface, search for icons or tabs that typically represent chat history, messages, or past conversations. This could be a "Chats" tab, a "History" icon (often a clock or list icon), or a specific conversation view.
Direct Access within Conversation: In some applications, simply navigating back to a specific chat session with an AI character will display the entire conversation history for that session.
Step 3: Filtering and Searching Conversation Logs
Once you've accessed the conversation logs, you'll often find a wealth of data. To make sense of it, you'll need to utilize the platform's filtering and searching capabilities.
Sub-heading 3.1: Setting the Time Frame
Most platforms will default to showing conversations from a recent period (e.g., "past 7 days").
Adjust the Date Range: Look for options like "Time Frame," "Date Range," or a calendar icon. Click on it to expand the options and select a custom date range, or predefined periods like "Today," "Last 30 Days," "This Month," etc. This is essential if you're looking for older interactions.
Sub-heading 3.2: Applying Filters for Specificity
Filters are your best friends for narrowing down your search. Common filters include:
User ID (UID) or Session ID (SID): If you have a specific user or conversation in mind, you can often search using their unique User ID (UID) or Session ID (SID). This is extremely helpful for pinpointing individual interactions.
Labels or Tags: Poly AI platforms often allow for labeling or tagging conversations based on intent, resolution status, or other custom criteria. Utilize these filters to find conversations related to specific topics or outcomes.
Automated Resolution Status: You might be able to filter by whether the conversation was resolved automatically by the AI or required human intervention. This is valuable for performance analysis.
Use Case: For AI agents trained on specific use cases, you can often filter conversations by the presumed use case.
Test Conversations: Many platforms allow you to exclude or include "test conversations" (interactions with the bot during development or testing) to keep your analysis clean.
Sub-heading 3.3: Searching for Keywords
Beyond filters, a search bar is usually available to look for specific keywords or phrases within the conversation content.
Enter Keywords: Type in relevant keywords that you remember from the conversation you're trying to find. This can be a product name, a specific query, or a key phrase used by the user or the AI.
Step 4: Reviewing Conversation Details
Once you've filtered down to the conversations you're interested in, it's time to dive into the details.
Sub-heading 4.1: Viewing Entire Conversation Transcripts
Click on a Conversation: In the list of conversation logs, click on the specific conversation you want to review. This will typically open a detailed view showing the full transcript of the interaction between the user and the Poly AI agent.
Scroll and Read: Carefully read through the conversation flow. Pay attention to the user's input, the AI's responses, and the overall progression of the interaction.
Sub-heading 4.2: Inspecting Message-Level Details
Many advanced Poly AI platforms offer the ability to dig deeper into individual messages.
Hover or Click on a Message: Hover your mouse over a specific message (either user or AI) or click on it to reveal more details.
Review Message Details: This might show information such as:
Message content: The exact text sent.
Presumed Use Case/Intent: What the AI understood the user's intention to be.
Reply Type: How the AI generated its response (e.g., from a predefined dialogue, a generative procedure).
Active Instructions/Persona: Any specific instructions or persona settings that influenced the AI's response.
Session Data: Actions, events, and other technical details that occurred during that specific message exchange.
Step 5: Utilizing Additional Features (If Available)
Some Poly AI platforms offer advanced features for managing and analyzing conversation history.
Flagging Conversations: If you encounter a particularly important or problematic conversation, look for a "Flag" icon. Flagging allows you to easily filter and revisit these conversations later for review or training purposes.
Adding Comments: Many platforms enable you to add internal comments to specific conversations or messages. This is great for team collaboration, noting issues, or providing feedback for AI improvement.
Downloading History: For auditing or offline analysis, some platforms provide an option to download the chat history (often in CSV or text format). Look for a "Download" or "Export" button.
Debugging Logs: For developers, there's often a "Logs" or "Debugger" tab within the conversation view. This provides a granular look at the AI's internal reasoning, decision-making, and any errors encountered during the conversation flow. This is invaluable for troubleshooting and optimizing AI performance.
By following these steps, you'll be well-equipped to navigate and understand your Poly AI interaction history. This empowers you to gain deeper insights into how your AI is performing and identify areas for improvement, ultimately leading to a more efficient and effective AI solution.
10 Related FAQ Questions:
How to export Poly AI conversation history?
To export Poly AI conversation history, navigate to the "Conversation Logs" or "Analytics" section in your Poly AI platform. Look for an "Export" or "Download" button, usually located near the filters or at the top of the conversation list. This will typically allow you to download the data in formats like CSV or TXT.
How to filter Poly AI logs by user ID?
To filter Poly AI logs by user ID, go to the "Conversation Logs" in your Poly AI platform. There should be a search bar or a filter option labeled "User ID" or "UID." Enter the specific user's ID into this field to display only their conversations.
How to view specific messages in a Poly AI conversation?
To view specific messages within a Poly AI conversation, first open the full transcript of the desired conversation from the "Conversation Logs." Once the conversation is open, you can usually hover over or click on individual messages to reveal detailed information about that specific exchange, including the message content, AI's intent, and more.
How to identify unresolved conversations in Poly AI history?
To identify unresolved conversations in Poly AI history, utilize the filtering options in your "Conversation Logs." Look for filters related to "Resolution Status," "Handover to Agent," or "Automated Resolution." By applying filters for "unresolved" or "transferred" conversations, you can quickly pinpoint interactions that the AI couldn't fully handle.
How to use Poly AI history for bot improvement?
To use Poly AI history for bot improvement, regularly review conversation logs, especially flagged or unresolved conversations. Analyze user queries that led to confusion or errors, identify common misunderstandings, and pinpoint areas where the AI's responses were inadequate. This data can then be used to refine training data, create new dialogue flows, or adjust AI intent recognition.
How to check Poly AI performance metrics from history?
To check Poly AI performance metrics from history, access the "Analytics" or "Dashboard" sections of your Poly AI platform. These sections often leverage conversation history to display key performance indicators (KPIs) such as resolution rates, containment rates, average conversation duration, common intents, and handover rates, providing an overview of your AI's effectiveness.
How to access Poly AI voice interaction history?
Accessing Poly AI voice interaction history typically involves the same "Conversation Logs" section as text-based interactions. The platform usually transcribes voice conversations into text logs, allowing you to review the spoken dialogue in a written format. Look for indicators if a conversation originated from a voice channel.
How to set a custom date range for Poly AI history?
To set a custom date range for Poly AI history, navigate to the "Conversation Logs" and locate the date filter option (often a dropdown or calendar icon). Click on it and choose "Custom Range" or similar, then select your desired start and end dates from the calendar or by manually entering them.
How to share Poly AI conversation logs with a team?
To share Poly AI conversation logs with a team, first access the specific conversation you wish to share. Many platforms provide a "Share" button or allow you to copy the unique URL of the conversation log. You can then share this URL with your team members, assuming they have the necessary access permissions to the Poly AI platform. Alternatively, you can download the conversation history and share the file.
How to troubleshoot Poly AI errors using conversation history?
To troubleshoot Poly AI errors using conversation history, open the problematic conversation in the "Conversation Logs." Look for detailed message insights, especially "Debugger" or "Logs" tabs, if available. These logs reveal the AI's internal processing, including detected intents, entities, decision paths, and any errors encountered, allowing developers to pinpoint and rectify the issue.