# FundApps Assistant (Intercom) **1. What is FundApps Assistant?** FundApps Assistant is an AI agent which is provided by Intercom, FundApps’ client support ticketing platform. FundApps Assistant acts as an intelligent virtual assistant which can provide immediate, automated assistance with: * *Answering Frequently Asked Questions (FAQs)*: Get answers to common queries about FundApps products, features, and platform functionality. * *Providing Troubleshooting Assistance*: FundApps Assistant can guide you through self-service troubleshooting steps for issues you may encounter. * *Help with Account Management*: Learn more about your account settings, features, and how to use the platform effectively. * *Recommending Resources*: It can direct you to relevant articles, guides, and support materials within the FundApps Help Centre. FundApps Assistant provides an automated service, available 24/7, for faster responses and general assistance. **2. How Does FundApps Assistant Work?** FundApps Assistant works by leveraging large language models (LLMs) to aggregate information from the FundApps Help Centre, analyse your request, and respond with relevant information, suggested actions, or links to further resources. While FundApps Assistant can assist with basic queries, it is not a replacement for human expert support. If your issue is complex or requires a more in-depth response, you will be directed to a human FundApps representative. **3. Data Privacy** We take your privacy seriously. When using FundApps Assistant, you should know the following: * *Data Collection*: Basic interaction data is collected, such as the questions you ask and the responses it provides. This helps improve the accuracy of the support provided. * *Data Sources*: FundApps Assistant leverages FundApps Help Centre articles and your Intercom support ticket information to answer your questions. It does not have access to your [confidential Client Data](https://policies.fundapps.co/client-portal/-LubIC9uIsME-_T0mNXu/fundapps-policies/risk-management/data-classification) or that of other clients, or the Intercom support ticket information of other clients, and cannot provide you with answers relating to this. * *Sensitive Data*: Do not share sensitive personal or financial information (e.g., position files, portfolio files, account passwords, financial details) by typing or uploading it into your conversation with FundApps Assistant. If your issue is complex or requires the sharing of these types of information, please reach out to a human FundApps representative. * *Intercom*: FundApps Assistant is provided by Intercom, FundApps’ client support ticketing platform. Our written agreement with Intercom includes a robust data processing agreement, and Intercom’s security accreditations have been assessed as meeting our high security standards, and will be reviewed annually. * *AI providers*: Intercom utilises Microsoft’s Azure OpenAI Service and Google to provide FundApps Assistant. Intercom does not allow these providers to use your data for training of the underlying AI model or otherwise improving the provider’s services. For more information, please read our [Privacy Policy](https://www.fundapps.co/privacy-policy). **4. Limitations of FundApps Assistant** While FundApps Assistant can provide immediate assistance, please note: * *Not a Replacement for Expert Support*: FundApps Assistant is designed to handle general inquiries and basic troubleshooting. It cannot replace a human representative for complex or personalised assistance. * *Response Accuracy*: FundApps Assistant provides answers based on available data from the FundApps Help Centre and may not always have the most up-to-date or comprehensive answers for every situation. If you need more precise or specific information, a representative will be happy to assist. * *Availability*: FundApps Assistant may be unavailable during technical maintenance or system updates. **5. How to Access FundApps Assistant** FundApps Assistant will be available to users from 15 July 2025, unless an admin user has contacted requesting to opt out their organisation. Admin users can opt out, and may opt back in, at any time. Individual users will need to acknowledge that they are aware they will be interacting with an AI agent, that they have read and understand this Policy Portal page on the FundApps Assistant, and consent to using FundApps Assistant when starting each conversation. If you ever feel that FundApps Assistant cannot fully resolve your issue, you can always escalate to a live support representative. **6. Consent to Use FundApps Assistant** If you consent to use FundApps Assistant, you acknowledge and agree to the following: * You are aware of how FundApps Assistant works and how your data will be used. * You understand the limitations of FundApps Assistant and acknowledge that it cannot replace human support for more complex issues. * You understand that you should not upload sensitive data, including position files and portfolio files, to FundApps Assistant or Intercom. * You can choose to stop using FundApps Assistant at any time by simply exiting the conversation or reaching out to our support team for assistance. **7. AI Specific Information from Intercom** Logging of AI decision-making logic and data sources Intercom states that its AI agent maintains logs that support incident investigation, performance analysis, and review of AI system behaviour. Depending on product configuration, Intercom can also provide visibility into the logic and content sources used by AI features, including source-related metadata available through conversation records and APIs. Bias mitigation, fair treatment, and ongoing assurance Intercom states that it takes active steps to reduce bias and support fair treatment of users. These measures include independent third-party assessment through AIUC-1 certification, quarterly technical testing, and safeguards intended to reduce unsafe, inappropriate, or unreliable outputs. The AI agent is also designed to ask clarifying questions or hand off to a human where confidence is low. Monitoring, measurement, and oversight of AI performance and model drift Intercom states that it has monitoring and measurement capabilities to oversee AI performance over time, including AI Agent reporting, real-time performance metrics, insights, and custom reporting. Intercom also states that third-party AI providers are contractually restricted from using customer inputs or outputs to train or improve their models, and that those inputs and outputs are not persistently stored by those providers. Explanations and supporting sources for AI outputs Intercom states that its AI can provide visibility into the logic and sources used by AI features, depending on the product in use. For the AI Agent, content sources can be exposed in conversation metadata through the API, and conversations can be searched using those attributes. This supports traceability of AI-generated responses back to the underlying support content or connected knowledge sources. Alignment with responsible and trustworthy AI standards Intercom states that its AI practices align with responsible AI expectations and are supported by governance, safety, and risk controls. This includes AIUC-1 certification, controls intended to reduce hallucinations and unsafe outputs, escalation to humans where required, and disclosure where end users are interacting with an AI agent. Alignment with NIST AI RMF and ISO 42001 principles Intercom states that its controls and governance are aligned with ISO/IEC 42001 and that this sits within a broader trust and compliance posture that also references certifications such as SOC 2 and ISO 27001. Intercom has also stated that it achieved ISO 42001 certification on 23 March 2026. Privacy-enhancing controls for customer data Intercom states that it applies privacy-enhancing controls when handling customer data used with Fin. Examples referenced include encryption in transit and at rest, logical separation between customers, payment card redaction in conversations, optional additional redaction through integrations, and secure Messenger transmission using JWT-based controls. Scalability for increased AI prompts and tokens Intercom states that its infrastructure is designed to scale dynamically to support increased usage, including higher volumes of AI prompts and tokens, subject to agreed commercial limits and applicable rate limits. Higher API limits may be requested, and configurable hard limits are available within the AI Agent. Testing, reliability, and scope of AI functionality Intercom describes the AI Agent as a retrieval-augmented generation solution designed to ground responses in approved knowledge and connected systems, helping to reduce hallucinations and improve reliability. Intercom also states that the AI agent has been subject to extensive independent testing through AIUC-1 and that customers can validate behaviour before go-live using simulations, batch testing, and manual previews. Controlled use of AI and protection against unauthorised access Intercom states that AI and machine learning are operated within a controlled environment with safeguards intended to prevent unauthorised access to customer data. These include managed access restrictions, device controls, privileged access management, contractual and technical restrictions on third-party AI providers, and data isolation measures designed to prevent cross-customer exposure. Use of customer data for model training or fine-tuning Intercom states that third-party AI providers are contractually restricted from using customer inputs or outputs to train or improve their models, and that those inputs and outputs are not persistently stored by those providers. In addition, FundApps has opted out of the use of anonymised and de-identified workspace data for Intercom’s own model training and fine-tuning. Transparency and accountability for AI models Intercom states that it supports transparency and accountability through documentation, disclosure of third-party providers via its [Subprocessors List](https://www.intercom.com/legal/subprocessors-list), and independent external validation through AIUC-1 certification. The supporting material also refers to system cards, security testing reports, and related model documentation reviewed as part of certification activity. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://policies.fundapps.co/client-portal/-LubIC9uIsME-_T0mNXu/legal-information/ai/fundapps-assistant-intercom.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.