To enable and configure the capabilities of Agent Assist, the process involves raising a request with the support team. Each feature has its own enablement process, and once enabled for the partner, basic sanity checks are conducted to ensure that the feature is functioning as expected and the model predictions are relevant. After the enablement and validation, the appropriate permissions need to be granted to users, and relevant training is to be provided to help them make effective use of the Agent Assist capabilities.

Certainly! With Agent Assist, you have the flexibility to configure and enable specific capabilities based on your customer service team's requirements.

Absolutely! The Agent Assist capabilities can be customized to meet specific business requirements. This customization allows you to make the functionality more usable and relevant for your organization. However, please note that for any major UI/UX changes, feasibility needs to be confirmed with the product team.

Yes, depending on the specific use cases and requirements, Agent Assist capabilities can be utilized for both text-based and voice-based customer interactions. However, it's important to note that certain capabilities, such as response compliance, paraphraser, smart responses, and compose, are specifically designed for text channels due to their nature and purpose.

Under Smart Assist, there are global and industry-specific models available that can be utilized. However, for more personalized and business-specific suggestions, it may be necessary to train and fine-tune the model using partner data. This process typically takes around 2-3 weeks to go live.

The training of AI models in Agent Assist varies depending on the feature, but the models are mostly trained using the historical customer care interactions between agents and customers.

If you come across any major flaws in the recommendations provided by Agent Assist, we recommend reaching out to the relevant product point of contact (POC) for that feature. You can share specific examples and provide your expected recommendations to address the issue.

For minor inaccuracies or incorrect/irrelevant Agent Assist suggestions, you can rely on our feedback support. This feature allows you to flag such suggestions that do not meet your requirements. By providing feedback on these instances, you contribute to the ongoing improvement of the model. As more feedback is collected, the model can be retrained to enhance its accuracy and relevance.

If you have any observations or encounter any issues with functionality that may impact the customer experience, we recommend reaching out to our support team at tickets@sprinklr.com and provide them with the necessary details. They will assist you in addressing the concerns and, if needed, can involve the product team to investigate any major issues or disruptions.

To gain insights into the performance of Agent Assist, you have the capability to create customized reporting dashboards. These dashboards allow you to monitor and measure various metrics, such as adoption rates and value metrics. By doing so, you can obtain a comprehensive overview of the usage and impact of the Agent Assist features.