Release Date: 20-Oct-2023
AppExchange Link: Install GPTfy
New Functionalities
- Cockpit: The Cockpit lets a user jump on to any major feature of GPTfy in a single click. This acts as a home page, where users can click a tile to land on a specific feature. This enables the user to connect with the team via GPTfy.ai.
- Retention: Retention in GPTfy is a powerful feature that enables users to manage AI Security Audit records by deleting them after a specified number of days from their creation. This not only reduces clutter but also enhances performance.
- Security Layer: This helps in safeguarding sensitive Salesforce data with multi-filter masking. This ensures AI processing is in compliance with InfoSec policies, regulatory and legal requirements.
- Merge Fields – Merge Fields is a powerful feature within GPTfy that empowers users to effortlessly integrate AI technology into their Salesforce experience. With this tool, users can copy and paste field names, including relationship formatting (e.g., Account.Owner.Name ), and leverage the potential of Related Fields. With GPTfy’s merge field feature, you can create dynamic and context-aware Text prompts.
- ROI Dashboard: Visualize how AI automation saves time and money. Highlight the return on your AI investment and the real benefits it brings.
- AI Insights Dashboard: This helps in understanding how users utilize prompts. This shows the usage by User, Profile, and Object.
Enhancements
- Prompts: The page layout has been enhanced. Now, we can create data context mapping from the Prompts page itself. Temperature(Controls the randomness of generated responses), Top P(setting Top P to x means that tokens with a cumulative probability above x are retained in the generated text), and max output token value (Specifies the maximum length of the generated output in terms of tokens, which includes words, punctuation, and other elements) can be added specific to each prompts.
- Security Audit: AI Response has been renamed as Security Audits. GPTfy generates responses based on the prompts provided by users. These responses are mapped inside a security audit record, whenever a response from AI is generated. This record helps in auditing the working of security layers in GPTfy.
- Data Context Mapping: Data Extraction Mapping has been renamed as Data Context Mapping. Configuring the Data Context Mapping is essential for effectively utilizing GPTfy. It involves mapping the fields of the objects that need to be sent for AI processing and scanning for any Personally Identifiable Information (PII) data.
- AI Models: Connection has been renamed as AI Model. The AI Model records specify the type of AI that will be invoked for generating responses. It establishes the link between GPTfy and the AI service.
- Prompt Catalog: Catalog has been renamed as Prompt Catalog. The Prompt Catalog feature in GPTfy allows you to assign specific business purpose prompts to particular profiles and activate them.
Known issue
- Links to the Account360 prompt are not working as links are case-sensitive; AI converts them to lowercase, which, in return, doesn’t give the expected results.
