# Features
1. **Supported AI Providers & Models**
iX Hello Employee is built on an **LLM-agnostic architecture** which enables the platform to support multiple AI providers simultaneously. This gives your organization access to frontier models from the world's leading AI companies, with the flexibility to choose the right model for each use case.
**AI Providers**
* **OpenAI -** The platform's primary provider and the source of the default model. OpenAI models are available via both direct API and Azure AI Foundry.
* **Azure OpenAI (Microsoft) -** OpenAI models deployed within Microsoft Azure, providing enterprise data residency, compliance controls, and regional hosting (US and EU).
* **Current model lineup**
* **GPT-5.4** *(default) -* The platform default for all new assistants — the latest and most capable general-purpose model.
* **GPT-5.4 Mini** - A lighter-weight variant of GPT-5.4, optimized for speed and cost efficiency.
* **GPT-4o / GPT-4o Mini** - Multimodal models supporting text, image, and audio inputs.
* **Anthropic (Claude) -** Claude models are integrated via a dedicated Anthropic SDK, offering an alternative known for strong reasoning and writing quality.
* **Current model lineup**
* **Claude Opus 4.7** - Anthropic's newest, most capable model for complex, nuanced tasks.
* **Claude Sonnet 4.6** - Balanced performance — strong reasoning and writing at moderate cost.
* **Claude Haiku 4.5** - Anthropic's fastest and most cost-effective model — ideal for high-volume, low-latency tasks.
* **Google Gemini** *(coming soon) -* Gemini models are currently in development and will be available in an upcoming release.
* **Media & Specialty Models**
* Whisper - Audio-to-text transcription.
* TTS-HD - High-definition text-to-speech voice generation.
**How Model Selection Works**
* **Users** can switch between available AI models directly within the chat interface — no settings menus or restarts required. Choose a lightweight model for quick questions and switch to a reasoning-heavy model for complex analysis, all within the same conversation.
* **Assistant builders and admins** can designate and lock a specific model per assistant, preventing end users from changing the selection. This ensures the right model is used for each workflow by design.
* **GPT-5.4** is the platform default and is automatically assigned to all new assistants unless an admin explicitly selects a different model.
**Tools Used Within Models**
**1. Tools**
* **Generate Text:**
* Create high-quality textual content based on prompts or themes. Use it for newsletters, articles, creative writing, training materials, and more — powered by your choice of LLM (GPT, Claude, or Gemini when available).
* **Generate Code:**
* Generate programming code snippets across various languages. Developers can reduce time spent on repetitive coding tasks, while the **Code Interpreter** tool allows assistants to execute, test, and return code and file outputs directly within the chat interface.
* **Generate Images:**
* Produce visuals from textual descriptions using dedicated image generation models (e.g., GPT Image 1 Mini), supporting creative processes in marketing, content creation, and media production.
* **Generate Voice:**
* Transform text into natural-sounding speech using TTS-HD (text-to-speech). Supports applications including audiobooks, virtual assistants, accessibility tools, and audio content creation. Speech-to-text transcription is also available via the Whisper model.
* **Generate Video** *(In Progress):*
* Develop video segments from text scripts. This feature is planned for marketing efforts, online courses, and webinars — allowing easy transformation of written content into engaging video formats.
**2. Data**
* **Chat with Knowledge Base (KB):**
* Interact directly with knowledge bases attached to virtual assistants using **Retrieval-Augmented Generation (RAG)**. The platform's Global RAG service indexes uploaded files and retrieves relevant knowledge chunks to generate accurate, grounded responses — supporting up to 10,000 files per assistant with 50+ data connectors. Supported file types include PDF, DOCX, PPTX, XLSX, CSV, JSON, TXT, and more.
* **Chat with Websites:**
* Gain information directly from websites through conversation. The assistant can browse and extract content from URLs provided during the chat interaction.
* **Chat with Files:**
* Upload and interact with files attached to the current conversation (e.g., CSV, Excel, images, PDFs). These files are available for analysis within the session but are not incorporated into the assistant's permanent knowledge base.
**3. Actions**
* **Search Online:**
* Conduct searches across the internet to quickly find diverse, up-to-date information. This accelerates research and insight gathering with inline citation badges linking back to source URLs.
* **Browse Websites:**
* Navigate to and extract content from specific web pages during a conversation, enabling real-time information retrieval from external sites.
* **Run Code (Code Interpreter):**
* Execute code snippets in a secure sandboxed environment directly within the chat. Supports data analysis, mathematical computation, file transformation, and automation — with results and generated files returned inline.
* **File Search:**
* Search through knowledge base files attached to the assistant to retrieve the most relevant information for a given query. Powered by the Global RAG integration for semantic, context-aware retrieval.
* **Analyze Images & Data:**
* Leverage AI to analyze uploaded images, charts, and structured data sources, providing actionable insights and metadata for enhanced decision-making.
* **Translate Text:**
* Translate content across 90+ languages, enabling global accessibility and cross-language communication within any conversation.
**Security**
* **Role-Based Access Control:**
* This security measure ensures that users have access only to the information relevant to their roles. It minimizes the risk of data breaches and ensures compliance with data governance policies.
* **Privacy Inc (PIP):**
* Adhering to privacy regulations, this feature protects user data in all interactions, safeguarding sensitive information and maintaining regulatory compliance.
* **Enterprise-grade Security:**
* Robust security measures are implemented to protect sensitive company data and transactions, assuring users of the integrity and confidentiality of their information.
* **Encrypted at Rest:**
* This security feature safeguards stored data from unauthorized access by utilizing established encryption standards, ensuring that information remains secure even when inactive.
5. **Assistants**
The assistants serve as intelligent tools designed to enhance productivity and efficiency in various tasks by leveraging artificial intelligence and data processing capabilities. These assistants simplify complex processes, automate repetitive tasks, and facilitate better decision-making by providing relevant insights and streamlined workflows. The use of natural language processing and machine learning enables these assistants to understand user inputs, produce outputs in an accessible format, and adapt to specific business needs.
---
# 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://docs.ixhello.com/ixhello/general/features.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.