Feature Request

Business Definition: Currently there is no way to see how AI Agent is performing. The only way users can start collecting AI Agent performance, is by logging the data within the workflow. For example, append a row in Google Sheet within the workflow if the AI Agent was able to pass the contact to the success branch, idle branch, or failure branch. Desired Outcome: Users can access a dashboard to see how AI Agent is performing. The metrics that can be captured are: Number of AI Agent interactions, breakdown by workflow name Number of AI Agent success / idle / failure conditions, breakdown by workflow name Metrics of steps initiated with its AI objective This form of analytics can be used to determine if the Knowlegde Source of a workflow step need to be re-worked, or if AI Agent is helping in automating conversations efficiently

Business problem: Currently, when the AI agent provides an incorrect response, there is no immediate way to correct and train the agent, which could perpetuate the same errors. Desired outcome: Implement a feature that enables direct corrections to AI agent responses within the platform, enhancing the agent's learning process and improving its future accuracy. Use case: Users would have the ability to flag incorrect responses and provide the correct ones, which the AI agent would then learn from to enhance future interactions.

Business Problem: In the current AI Agent, there is only one objective: Answer questions. The objective is simple, but the scope is adaptable and opens multiple avenue of possibilities. In this case, because the objectives can be subjective, it is difficult for users to configure them effectively. User struggle to set up these agents for specific tasks, despite the agents' potential to handle a wide range of subjective tasks. Desired Outcome: Develop an AI setup bot for users to converse with while setting up AI Agent of their choice. This AI setup bot aims to: Understand users' goals with AI Agent. Recommend objectives for the AI agent. Guides users in configuring the AI agent. Engages in detailed conversations to gather precise requirements. Facilitates additional use cases as needed. This approach aims to save time, lower the entry barrier for users, and provide clear guidance during the AI agent setup process, thereby enhancing the potential applications of AI in various business contexts which additionally allow users to create their own objectives. Use case (Capture, Convert, Retain): This AI Agent feature plays a foundational role in any type of phases in a sales cycle: Reduces time-to-value by accelerating setup Increases successful AI deployments Lowers user frustration and churn due to misconfiguration Encourages experimentation with more advanced or custom use cases

Business problem: Currently, AI Agent does not have the capability to process images. This feature would help platform users handle objective queries where the contact sends an image, AI Agent will not be able to process it answer based on the image that the contact is enquiring about. Desired outcome: Allow AI to process images and respond based on the image. Specific use cases: Identify brands of items of the image and provide a response on said brand / product Identify issues from images and provide a response based on knowledge sources (i.e. image of a malfunctioning device of a specific brand) Process an image of a form, receipt, etc Use case (Capture, Convert, Retain): This feature spans multiple phases: Capture: Visually engage users early in the journey—think visual search or product discovery via images. Convert: Streamline image-based product matching and upselling in chat. Retain: Use visual input to reduce back-and-forth and solve problems faster.

Business Problem: Currently, the AI Agent: Collect Information feature asks a confirmation question after gathering details such as name, email address, and phone number. While this ensures accuracy, it often results in drop-offs when Contacts or Customers fail to respond, causing the AI to withhold the collected information from being passed to respond.io , and proceed to ELSE branch. Desired Outcome: Introduce an option to disable the "Confirmation" step in the AI Agent. This will allow the AI to automatically pass collected information to respond.io for data storage even if the Contact did not confirm with AI, reducing drop-offs and ensuring seamless data capture.

Business Problem: Currently, there's no way to test an AI Agent before deploying it in a live environment. This lack of a sandbox or preview mode creates friction during setup, increases the risk of misconfigured responses, and slows down iteration. Teams are forced to activate the agent to see how it behaves, which can lead to awkward customer experiences or internal confusion. Desired Outcome: A Test Mode for AI Agents that allows admins to simulate conversations, preview AI behavior, and fine-tune logic — all without affecting live customers. The test interface should reflect how the AI will respond in real scenarios. Use Case (Build + QA) Before Launch : Managers and admins can preview how the AI handles FAQs, product inquiries, or assignment flows. While Tuning : Teams can safely test tweaks to tone, logic, or integrations without needing to go live first. With a Test AI Agent mode, teams can confidently build, experiment, and perfect their AI experiences — without flying blind or breaking things in production 🔧🧪✅

Business problem: AI Agent currently can crawl Knowledge Sources defined for AI Agent. However, it's hard for platform users to identify which part or snippet of the Knowledge Source that the AI Agent is referencing. Without a log to tell the user what Knowledge Source it is referencing, platform users can not identify clearly which part of the Knowledge Source they should improve or change so that the AI Agent can provide more accurate answers Desired outcome: When an AI Agent replies, there should be a log to tell the platform user what part of the Knowledge Source is being crawled. This can be an activity log or an information call-out for users to refer with a "snippet" or source of the Knowledge Source

Business Problem: Our platform currently lacks a personalized product recommendation feature. This absence limits our ability to guide customers towards products that best suit their needs and preferences, potentially leading to missed sales opportunities and reduced customer satisfaction. Desired Outcome: Implement a new AI Agent objective that suggest relevant products based on Contact's conversation. This objective should improve user experience by offering tailored suggestions, increasing the likelihood of purchases, enhancing customer engagement and able to capturing intent after recommending products. Use case (Convert): This feature clearly shines in the Convert phase. You're already engaging the customer—they’ve reached out or responded—and now it's time to seal the deal. The AI doesn’t just inform; it persuades. It uses the context of the chat to: Recognize buying signals Suggest products that align with the customer’s use case Help the customer make a confident, informed buying decision So, you're not just solving for engagement—you’re solving for revenue.

Business problem: AI Agent Builder can only generate the functional configuration of the AI Agent Step. It does not help generator the AI Persona, which can be customised manually after applying the configuration for AI Agent step. Desired outcome: Allow AI Agent Builder to also ask for the desired AI Persona. Then, once the user is satisfied with the generated AI Persona, the user can confirm with the AI Agent Builder and apply the AI Persona on the AI Agent Step so as to configure a well written persona that the AI Agent Step can follow.

Business problem: Currently, AI Agent does not have any restriction on the number of responses. So as long as the contact is conversing, and the AI Agent's objective has not been achieved, AI Agent will continue to converse with the user perpetually. This might introduce a risk where the AI Agent might use up AI credits Desired outcome: Platform users should be able to restrict AI Agent from generating too much responses. For example: After 10 responses, AI Agent can end the step and pass it to a particular branch. This gives more control for the user in configuring AI Agent's responses. And in conjunction, provide the platform user more control of their AI credits usage.