We are thrilled to introduce a game-changing addition to your customer insights toolkit - the brand-new OI-Q! Powered by cutting-edge generative AI technology, this innovative solution promises to revolutionize the way you gain insights from your customer contacts.
OI-Q is available as an add-on to your Operative Intelligence platform. If your organization doesn’t yet have OI-Q and you’re interested in learning more about how to use generative AI for your customer demand insights, reach out to our team today to schedule a personalized demonstration and learn how OI-Q can supercharge your customer service operations.
Key features of OI-Q:
User-friendly summaries: Say goodbye to information overload! OI-Q generates concise and user-friendly summaries of your customer interactions. You'll be able to quickly grasp the essence of each conversation, saving you valuable time and ensuring that nothing important slips through the cracks.
Compliance checking: We understand the critical importance of adhering to compliance guidelines. With OI-Q, you can effortlessly check whether your agents have followed the industry standard security and identity checking procedures, reducing risks and ensuring regulatory alignment.
Agent feedback: Enhance the performance of your call center agents with actionable feedback. OI-Q not only highlights areas where your agents excel but also provides opportunities for improvement, enabling targeted training and skill development.
How to use OI-Q:
From any of the insights pages in Operative Intelligence, click “View data sample” which is located on the right hand side near the top of the page.
From the data sample page, click “OI-Q” at the top of the page.
OI-Q will load in a new window. Select an interaction by clicking on one of the transcripts on the left hand side of the page.
To see more details about the interaction, click the 'Summary Stats' button on the right hand side panel.
You will now see more of the transcript along with data points about this interaction.
Provide OI-Q with a prompt. You can select a predefined prompt or type your own.
Predefined prompts:
Summarize: provides a summary of the interaction in 150 words or less.
Customer satisfaction: OI-Q will determine if it thinks the customer is satisfied, dissatisfied or neutral based on the transcript and provide the reasons for which option is selected.
Highlights and opportunities: provides a summary of what the agent did well or could do better on the interaction.
(Note: OI-Q does not know the internal processes and business requirements, so interprets highlights and opportunities as best it can, but may not always be applicable to your organization)
Compliance: checks if the customer was asked for three points of identification in the interaction.
(Note: transcripts in OI are redacted so it will not see the answers, only whether or not the questions were asked)
Resolution: OI-Q will try to ascertain if it thinks the customer was satisfied with the outcome of the interaction and provide reasons why (or why not).
Questions: OI-Q will identify which questions were asked in the selected transcript.
Predefined prompt example:
Free typed prompts:
Note:
It's important to be aware that generative AI, while powerful, can occasionally generate inaccurate or speculative information.
To mitigate this, we recommend being very specific with prompts when using
OI-Q. Providing clear and well-defined instructions ensures that the AI delivers more precise and reliable results.
You can ask OI-Q anything about the transcript you like, however to get the best answers it is best practice to consider the following points:
Specific to the selected interaction - OI-Q is referencing the specific selected transcript so will only infer what it can from the information provided.
Be clear and specific - Start with a clear and specific instruction or question. Avoid vague or overly broad prompts that could lead to ambiguous responses.
Define the context - Provide context or background information if necessary.
Ask direct questions - If you want a direct answer, pose your prompt as a clear and concise question.
Define the format you would like a response to be in - if you want dot points, or the response to be under a certain number of words, include this in your instructions.
Free typed prompt example:
OI-Q Troubleshooting
Transcript is too short
Occasionally an interaction transcript is too short for OI-Q to meaningfully analyse. Typically these transcripts occur when a call cuts out very early on, or when the caller or agent cannot hear each other. This may also happen with very brief emails or web chat conversations.
When this occurs, OI-Q will ask you to choose a longer transcript to analyse.
Transcript is too long
Generative AI models are limited in the amount of 'tokens' or context they can process at a time. Very rarely an interaction transcript may exceed this limit due to being too long. If this occurs on a very long transcript, OI-Q will provide an error message asking you to select a shorter one.
