Feature Request

Description This feature will ensure that managed models from the EU and US comply with our regional policies and eliminate the need for users to provide their own API keys for accessing these models. This enhancement will streamline the user experience while maintaining compliance with legal and ethical standards. User Story As a user, I want managed models to comply with our regional policies and not require an API key so that I can easily and legally access the models without additional setup. Problem Statement Users face challenges in accessing managed models due to compliance issues and the requirement of an API key. This feature addresses these problems by ensuring compliance and simplifying access. Acceptance Criteria Managed models from the EU and US show compliance status related to regional policies. Users can access managed models directly without needing to enter an API key. The UI indicates that no API key is required and displays compliance information clearly.

Description This feature will enable AICamp to halt message passing at the gateway level when specific keywords are detected. This is crucial for protecting sensitive information and ensuring compliance with data privacy regulations. User Story As a CISO, I want AICamp to stop data sharing when certain keywords are detected, so that we can safeguard sensitive information and maintain compliance. Problem Statement Currently, there is a risk of sharing sensitive data with models, which can lead to potential privacy breaches and non-compliance with regulations. Acceptance Criteria Users can define a list of keywords that trigger the halt in data sharing. The gateway successfully blocks message passing when any of the specified keywords are detected. An alert is generated and logged whenever data sharing is stopped due to keyword detection.

## Description This feature will provide insights into the number of active users on a weekly and monthly basis. It is crucial for understanding user engagement trends and making informed decisions about AI Adoption and marketing strategies to become mode AI productive ## Problem Statement Currently, there is no clear visibility into user engagement over time, making it difficult to identify trends and measure the effectiveness of improvements or marketing efforts. ## Acceptance Criteria Users can view weekly active user statistics. Users can view monthly active user statistics.

Description This feature aims to improve the user interface for displaying reasoning information behind AI responses. By providing transparent reasoning, users can better understand the logic and decision-making process of the AI, leading to increased trust and effective use of the technology. User Story As a user, I want the reasoning behind AI responses to be displayed in the UI so that I can understand how conclusions were reached. Problem Statement Currently, users are unable to see the reasoning behind AI responses, which can lead to confusion and a lack of trust in the system. This feature addresses the need for transparency in AI decision-making. Acceptance Criteria Users can view reasoning information alongside AI responses. The UI clearly differentiates between the response and the reasoning section. Reasoning information is derived directly from the thinking models used in generating the response.

The AICamp will provide a library of prebuilt prompts for various tasks, allowing users to explore and reference them for their own creations. This feature enhances user experience by offering ready-made solutions that can be adapted or directly utilized. User Story As a user, I want access to a prebuilt prompt library so that I can easily find and use prompts for my tasks. Problem Statement Users currently lack a centralized resource for prompt ideas, leading to inefficiencies and reliance on trial-and-error when crafting their own prompts. Acceptance Criteria Users can browse a library of categorized prebuilt prompts. Users can select a prompt to view its details and usage examples. Users can save or copy prompts to their personal workspace for further customization.

Description The AICamp Assistant Library will offer a diverse collection of virtual assistants designed for various use cases, allowing users to experiment with different options or reference them to create their own customized solutions. This feature is crucial as it empowers users to find the most effective tools for their specific needs and enhances their overall experience. User Story As a user, I want to access an Assistant Library with various use cases so that I can try out different assistants or use them as a reference to build my own. Problem Statement Users often struggle to find suitable virtual assistants for their specific tasks, leading to inefficiencies and missed opportunities. This feature addresses that gap by providing a comprehensive library of options tailored to diverse use cases. Acceptance Criteria Users can browse a library of virtual assistants categorized by different use cases. Users can try out any assistant directly within the platform. Users can access reference materials to help them create their own custom assistants based on the available options.

Description The current BOYK Model UI presents a congested view due to the extensive list of providers and their corresponding models. Streamlining the UI will enhance user experience by making navigation easier and allowing users to focus on relevant models and permissions. User Story As a user, I want an improved BOYK Model UI so that I can easily navigate and manage models without feeling overwhelmed. Problem Statement The existing user interface is cluttered with a long list of providers and models, making it difficult for users to find and manage the necessary features efficiently. Acceptance Criteria Users can filter models by provider for easier access. The UI displays model permissions in a more compact and organized manner. Navigation between different providers is seamless and intuitive.

Description The Assistant Knowledgebase feature allows users to upload knowledge files that the assistant can reference to provide more accurate and contextually relevant answers. This functionality enhances the assistant's ability to deliver tailored responses based on the specific information users provide. User Story As a user, I want to upload knowledge files to the assistant so that it can give me answers based on my specific context and requirements. Problem Statement Currently, the assistant may lack context-specific knowledge, leading to generic responses that do not fully address user needs. This feature aims to bridge that gap by allowing users to provide tailored information. Benefits Improves the relevance and accuracy of assistant responses. Reduces the time users spend searching for information independently. Enhances user satisfaction by providing context-aware answers. Increases the assistant's utility for specialized tasks or inquiries. Encourages user engagement by allowing customization of the assistant's knowledge base. Acceptance Criteria Users can successfully upload various file types (e.g., PDF, DOCX, TXT). The assistant can utilize the uploaded files to answer user queries accurately. Users can view, edit, and delete their uploaded knowledge files.

I propose that the assistant should have a default model setting, while allowing users the option to switch models as needed. Some prompts are specifically crafted for certain models, making this feature beneficial.

Moving the prompt library closer to the chat input (or even as an icon/button within the input) would be a quality of life improvement in terms of UX. Also, allowing the {{variables}} within a prompt to be automatically highlighted or selected for replacement when they are added to the chat, or press the tab key to move to the next variable in prompts with more than one.