# FAQs

How does Purple Fabric integrate with our existing technology stack and infrastructure?
Answer: Purple Fabric offers API interfaces and database interfaces for connectivity with various systems. It provides CI/CD integration, allowing it to fit into existing DevOps processes.
Can we deploy Purple Fabric on-premises, in the cloud, or in a hybrid environment?
Answer: Purple Fabric can be deployed within your own Virtual Private Cloud (VPC), or as a Cloud-Hosted managed PaaS.
What kind of scalability does Purple Fabric offer to handle our enterprise-wide AI needs?
Answer: Purple Fabric offers scalability through built-in DevOps, MLOps, and LLMOps capabilities. It includes features like zero pod, load-based, and time-based scaling. When customers opt for PaaS, Intellect takes care of scaling based on the load analysis. When customers opt for Private VPC, there will be shared responsibility.
How does Purple Fabric ensure compatibility with different LLM providers and models?
Answer: Through the LLM Optimization Hub (Model Hub), Purple Fabric 3.0 is fully model- and provider-agnostic. You can onboard out-of-the-box enteprise LLMs (Azure OpenAI, Bedrock Anthropic Claude, Google Cloud, Deepseek etc.) or bring your own keys for private/self-hosted models. The hub manages credentials, rate limits, and lets you switch or benchmark models per-agent without code changes.
What is the platform's approach to MLOps and LLMOps?
Answer: Purple Fabric has built-in MLOps and LLMOps capabilities, offering Infrastructure Management, Continuous Integration and Continuous Deployment (CI/CD), Monitoring, Logging, Data Management, Compliance, Security, Collaboration, Communication, Data lineage, Benchmarking, Scalability and availability
How does Purple Fabric address data privacy and security concerns when handling sensitive enterprise data?
Answer: Security is enforced at every layer with Enterprise Governance (PF Govern): Purple Fabric address through Data Encryption, Access Control and Identity Management, Secure Data Storage, Network Security, Monitoring and Logging, Compliance and Regulatory Adherence, Vulnerability Management, Incident Response and Recovery, Secure Development Practices. From an AI focussed security perspective, Purple Fabric addresses this through deployment within Virtual Private Cloud (VPC), not sharing data for model training, built-in PII Masking, Prompt Defense, Toxicity Defense, and Zero Tolerance Policy to External Access.
What is the commercial model for Purple Fabric?
Answer: Purple Fabric comes with a ZERO License fee; being deployed at the customer cloud, we would charge for the Support subscription services. The support charges would be on a per core per year model; with minimum core to be deployed being 60 cores. The support services would include the L3/L4 support on the platform; any other nature of support would be priced additionally. The customer would have to subscribe for LLM, MongoDB and other services at an additional cost directly with the OEM. A detailed commercial guide would be shared shortly
How does the platform support our digital transformation initiatives across different departments?
Answer: Purple Fabric 3.0 lets each department spin up Enterprise Digital Experts—AI agents tailored to their persona (e.g., Finance, Risk, HR)—using a low-code/no-code interface, pre-built templates, and the Flow Designer. Each function can have walled workspaces. All knowledge bases and agents in a workspace are centrally hosted in the Enterprise Knowledge Garden, so assets are reusable, composable, and instantly available to any function.
Can Purple Fabric integrate with our existing data management and business intelligence tools?
Answer: Purple Fabric provides APis and out of the box connectors to integrate with existing data management and business intelligence tools. Whatever Data-as-a-service platforms the clients have invested in, Purple Fabric's Agents are able to connect to and consume on demand.
What kind of ROI can we expect from implementing Purple Fabric, and how is it measured?
Answer: ROI stems from three pillars: 1) Lower cost-to-serve through multi-model optimization; 2) Faster time-to-value via pre-built connectors, templates, and Socratic workflows; 3) Improved decision accuracy (95%+ decision-grade outputs).
How can Purple Fabric accelerate our AI adoption across different business units?
Answer: By empowering subject-matter experts to build Digital Experts in hours, not months, using the GenAI Studio and pre-built templates. The Enterprise Knowledge Garden and Flow Designer eliminate data-prep bottlenecks, while cross-agent interoperability means new use cases can leverage existing knowledge bases and agents instantly.
What kind of support and training does Purple Fabric offer to help our teams build and deploy AI agents?
Answer: We provide a Purple Fabric Academy (on-demand courses, tutorials, workshops), and periodic Foundation Workshops and bootcamps. Our Help Center includes API docs and a Developer Community forum for peer support.
How does the platform enable non-technical users to leverage AI capabilities?
Answer: Non-technical users use the GenAI Studio and Flow Designer—drag-and-drop UIs for authoring knowledge pipelines, agent workflows, and conversational dialogs. No code is required to ingest documents, define roles/goals for agents, or deploy conversational or automation agents.
Can you provide examples of successful digital transformation use cases implemented with Purple Fabric?
Answer: One of the world's Largest Sovereign National Wealth Fund uses EKG + ESG Analyst Agents for risk & sustainability scoring; one of the largest UK Wealth Managers productionized multi-agent complaints investigation system. Refer to the solutions section of the Purple Fabric Website for more use cases.
How does Purple Fabric support continuous improvement and iteration of AI models and agents?
Answer: Continuous improvement is supported through built-in MLOps and LLMOps capabilities, Benchmarking Studio, and continued monitoring of published assets.
How does Purple Fabric fit into our overall enterprise architecture strategy?
Answer: Purple Fabric integrates seamlessly with existing systems, offering an API-first approach and SDK for deeper integration. It ensures unified data views, robust governance, security, and lifecycle management. With scalable, real-time data processing and enhanced team collaboration, it strengthens AI capabilities and secures efficient data management.
What APIs and SDKs does Purple Fabric offer for integration with our existing systems?
Answer: Open API standard documentation is available. SDK documentation will be available soon.
How does the platform handle data governance and lineage across different AI use cases?
Answer: Data governance and lineage are handled through advanced RAG with citations and references. Additionally TRACES hosts a log of the agentic throught process and collaboration, creating truly explainable AI with a robust audit trail.
Can Purple Fabric work with both structured and unstructured data from various sources in our organization?
Answer: Yes, Purple Fabric works with both structured and unstructured data from various sources. Purple Fabric supports ingesting data from the existing structured data and enterprise application landscape of our clients on demand as needed. The Enterprise Knowledge Garden ingests unstructured (docs, emails, PDFs, images) via native connectors for emails, web crawler, file sources, drive, and document storage. All content is enriched, chunked, embedded, and stored in a unified vector Database for AI-ready access. This is is Knowledge-as-a-service
How does the platform support version control and change management for AI models and agents?
Answer: Version control and change management are supported through built-in versioning capabilities features as part of its MLOps/LLMOps offerings. Historical snapshots of agents and prompts are maintained for rollback and compliance.
How does Purple Fabric help streamline our AI development and deployment processes?
Answer: By combining low-code/no-code tooling, automated data-prep in the EKG, and one-click agent publishing from the GenAI Studio, Purple Fabric reduces Dev-to-Prod cycles from months to days. Built-in MLOps/LLMOps automate testing, deployment, and monitoring
What kind of monitoring and analytics tools does the platform offer to track AI performance and ROI?
Answer: We surface real-time dashboards in PF Govern: agent usage, latency, cost per transaction, error rates etc.
How can we use Purple Fabric to automate and optimize our existing business processes?
Answer: You can create AI agents for specific tasks, use the Flow designer for orchestrating complex workflows, and molecular agent teams for intelligent reasoning and orchestration, all while leverage integration capabilities to connect with existing enterprise systems.
What benchmarking capabilities does the platform offer to ensure we're getting the best results from our AI initiatives?
Answer: The Benchmarking Studio lets you run comparative tests across Agents, prompts, and settings; visualize cost vs. latency vs. accuracy
How does Purple Fabric support collaboration between different teams in developing AI solutions?
Answer: Collaboration is supported through shared workflows and tools, human-in-the-loop capabilities, and a unified platform for different roles to work together. The platform has robust user management capabilities with personas having tiered access to platform features.
How does Purple Fabric ensure the explainability and transparency of AI decisions?
Answer: Explainability and transparency are ensured through advanced RAG with citations and references. Additionally TRACES hosts a log of the agentic thought process and collaboration, creating truly explainable AI with a robust audit trail.
What kind of support and maintenance does Purple Fabric offer post-implementation?
Answer: A dedicated customer support team to handle L1/L2 level incidents, dedicated customer success management, help and resource centre, and an academy for self-learning.
Can you provide case studies or references from other enterprises that have successfully implemented Purple Fabric?
Answer: Yes—clients span banking, insurance, asset management, wealth management, and finance & accounting. Key case studies (NBIM, St James’s Place, Global Bank RM Copilot) are on our website.
How does Purple Fabric address the challenge of data quality and consistency when building AI agents?
Answer: Purple Fabric addresses this through data preprocessing services, data quality analysis and enhancement features, and Advanced RAG capabilities for improved data retrieval and use in AI models. It is also multimodal, being able to handle data in different formats. The Flow Designer can invoke custom cleaning or enrichment steps, and the EKG stores metadata (source, date, confidence) to ensure consistency.
How do you protect my data within the platform?
Answer: We protect your data with robust encryption, access controls, and regular security audits. Our platform adheres to industry standards and regulatory requirements, ensuring data privacy and security throughout its lifecycle.
Do you share my data with public LLMs?
Answer: No. Your data is never shared with public LLM training datasets. All inference runs occur in your private tenancy or VPC; We use only enterprise-grade LLMs hosted on enterprise platforms such as Azure, AWS, or Google Cloud. Additionally, the LLM Model Optimization Hub allows you to bring your own LLM keys.
What measures are in place to prevent unauthorized access to my data?
Answer: We have robust Identity and Access Management (IAM) controls, logging, and secret management in place to prevent unauthorized access.
What happens if there is an issue with the service?
Answer: For Intellect hosted Platform, We have comprehensive monitoring and versioning systems in place. In case of any service disruptions, our orchestration and model monitoring services will ensure a swift resolution.
How do you ensure the AI models provide accurate responses?
Answer: We ensure accuracy by benchmarking the AI models we consume, providing users with detailed performance metrics. Since it's a self-service platform, accuracy also depends on user inputs and data quality. The full featured Benchmarking Studio present in the LLM Model Optimization Hub platform helps you optimize prompts, choose the right LLMs, and manage data effectively to optimize your outcomes for accuracy, cost, and latency.
Can I use my own AI models with your platform?
Answer: The LLM Model Optimization Hub allows you to bring your own keys for LLMs
Is there support for connecting to other cloud services like Azure?
Answer: Yes, we offer a Private Link to Azure OpenAI, allowing secure connections to other cloud services.
What happens to my input data when I use the platform?
Answer: Your data, including prompts and outputs, is not shared with other customers, cloud and model providers or any third parties.
Does OpenAI, Anthropic, Google, or Deepseek have access to my data?
Answer: No, all models available on the platform are hosted by managed services such as Azure OpenAI (Protected by Azure data protection policy), AWS Bedrock Claude (Protected by AWS data protection policy), or Google Cloud Gemini (Protected by GCP privacy policy) and they do not have access to your data or prompts.
Will my usage of the service contribute to improving OpenAI's or other providers models?
Answer: No, your data is not used to improve OpenAI or any other provider models. The service operates independently to ensure your data privacy.
Can Microsoft, AWS, or Google use my data to enhance its products or services?
Answer: Microsoft, AWS, or Google do not use your data to improve any of their products or services. These services are designed to respect your data's confidentiality.
Do the AI models learn and improve from my inputs over time?
Answer: The models are stateless and do not learn or improve automatically from your inputs unless you choose to fine-tune models with your training data.
If I fine-tune a model, who has access to it?
Answer: Any fine-tuned models you create are available exclusively for your use and are not accessible to others.
Is the Azure OpenAI service separate from OpenAI's services?
Answer: Yes, the Azure OpenAI Service is fully controlled by Microsoft, hosted on Azure, and does not interact with services operated by OpenAI.
What is the relationship between Azure OpenAI and OpenAI's ChatGPT or API?
Answer: Although Azure OpenAI hosts OpenAI models, the service operates independently and does not interact with ChatGPT or the OpenAI API.
How secure is my data within the Azure OpenAI service?
Answer: The service is designed with state-of-the-art security measures, ensuring that your data remains private and secure within Microsoft's Azure environment.
Can I be confident that my data will not be used to train other AI models?
Answer: Yes, you can be confident that your data will not be used to train or improve any AI models within or outside of the Azure OpenAI service.
What about the models on GCP? Or the Anthropic Models on Bedrock?
Answer: All of the above apply
How does Purple Fabric's protect our data and AI assets?
Answer: Purple Fabric protects your data and AI assets with advanced encryption, secure network and application with strict access controls, continuous monitoring, and compliance with industry standards, ensuring robust security throughout the AI lifecycle.
What measures are in place to prevent data leakage or unauthorized access to our information?
Answer: Measures include deployment within your VPC, Zero Tolerance Policy to External Access, and ensuring data is never shared or used for model training.
How does the platform handle PII (Personally Identifiable Information) masking and data encryption?
Answer: The platform handles PII masking through built-in capabilities.
How does the platform ensure the ethical use of AI and prevent potential misuse?
Answer: We embed customizable protections in every module (PF Govern) via policy-based filters, toxicity controls, and human-in-the-loop approval workflows. Audit trails ensure AI outputs remain fair, compliant, and aligned with your corporate values.
What is the Enterprise Knowledge Garden (EKG) and how does it work?
Answer: The EKG is Purple Fabric’s unified knowledge-as-a-service layer. It ingests semi-structured, and unstructured sources via native connectors (DBs, files, APIs, web crawlers), then applies pre-processing, chunking, metadata tagging, embedding, and vectorization pipelines. The result is a vector store (vector DB) that powers all downstream agents - Knowledge as a service.
Does the EKG handle structured data?
Answer: Enterprises have spent decades investing in their structured data application Landscape (Data-as-a-service). Purple Fabric can read these on demand without storing them in the EKG. EKG is more for semi-structured and unstructured data.
Can the EKG handle images, infographics, and tables, and handwriting?
Answer: Yes, the EKG is able to handle unstructured documents with images, infographics, and tables, and handwriting. It is able to embed these into a vector database, making it searchable through RAG.
What is an Enterprise Digital Expert (EDE)?
Answer: An EDE is a multi-agent AI persona composed of one or more atomic (single-skill), molecular (multi-skill + tools), or compound (complex workflows) agents. Each EDE is defined by a “role + goals + outcomes” template, grounded in one or more EKG knowledge bases and/or structured data from the enterprise application and data landscape. It is powered by one or more LLM models via the Model Hub.
How does the Flow Designer orchestrate multi-agent workflows?
Answer: The Flow Designer is a drag-and-drop canvas where you define event triggers (webhooks, schedules, file drops), branch logic, and agent invocations. It automatically wires inputs/outputs between agents and external APIs - letting you build end-to-end automations with intelligence embedded without code.
What does the LLM Optimization Hub (Model Hub) do?
Answer: The Model Hub centralizes onboarding, benchmarking, governance, and lifecycle management of all LLMs. Using the Benchmarking Studio, you test models+prompts against your use case on cost, latency, and accuracy.
What governance features does PF Govern provide?
Answer: PF Govern offers real-time dashboards for agent performance, cost, and usage; detailed Trace Viewer logs of every decision step with source citations; role-based access control; policy/guardrail enforcement; and audit-grade reporting on data lineage, retention, and compliance across all agents and knowledge bases.
Do you support MCP (Model Context Protocol)?
Answer: The Purple Fabric team is evaluating MCPs at the moment in terms of enterprise use. MCPs are effectively wrappers around APIs, and Purple Fabric supports APIs as tools for agents, which means you can replicate the functionality.
Do you support the A2A Protocol (Agent to agent)?
Answer: The PF team is evaluating the A2A protocol - acknowledging that even companies supporting A2A don't have seamless integration of agents that can collaborate across platforms. Purple Fabric is built to be open where every flow and Agent can be published as an API, which means that it can theoretically already interact with any external agents as a base.
What is Business Impact AI?
Answer: Our philosophy lies in our understanding of business impact AI-aligning AI outcomes with enterprise imperatives, delivering measurable business value across:
Reimagine Customer Experience:
Deliver hyper-personalized, AI-augmented services that elevate engagement and loyalty.
Reimagine Products:
Unlock new growth opportunities through product innovation and intelligent data-driven insights.
Reimagine Operations:
Lower costs and accelerate cycle times through AI-driven process optimization and automation.
Reimagine Compliance:
Embed security, governance, and auditability by design - meeting regulatory demands with confidence.