Skip to main content
This section provides an overview of three essential features for customizing your PERSO Interactive sessions: Background Images, Prompts, and Documents.

​
Background Images

Background images allow you to customize the visual backdrop of your AI Human video sessions, creating a more personalized and branded experience. Set Background Images here →

​
Key Features

  • Upload Custom Images: Add your own background images to use in sessions
  • Flexible Aspect Ratios: Support for both landscape (16:9) and portrait (9:16) formats
  • High Quality: Recommended resolution of FHD (1920 x 1080) or higher
  • Easy Management: View, edit, and delete background images from a centralized list

​
How It Works

  1. Upload an image with a descriptive title
  2. The image is stored and made available for your organization
  3. When creating a session, specify the Background Image ID
  4. The image appears behind the AI Human during video conversations

​
Use Cases

  • Branded environments for customer interactions
  • Context-specific backgrounds (office, outdoor, etc.)
  • Custom visual themes for different use cases

​
Quick Start

  1. Navigate to Resources → Background Images
  2. Click “Upload” to add a new image
  3. Provide a title and select your image file
  4. Use the Background Image ID when creating sessions

​
Prompts

Prompts define the personality, behavior, and knowledge scope of your AI Human by configuring the system instructions sent to the Large Language Model (LLM). Set Prompts here →

​
Key Features

  • System Prompt Configuration: Define how the AI should behave and respond
  • Retrieval-Augmented Generation (RAG): Enable document-based responses
  • Intro Messages: Set automatic greetings when users connect
  • Multiple Prompts: Create different prompts for different use cases

​
How It Works

  1. Create a prompt with system instructions
  2. Optionally enable RAG to include document context
  3. Add an intro message for automatic greetings
  4. Select the prompt when creating a session
  5. The LLM uses your prompt to generate responses

​
Core Components

System Prompt: The main instruction that defines:
  • AI personality and tone
  • Response scope and boundaries
  • Behavioral guidelines
  • Example interactions
RAG (Require Document): When enabled:
  • System prompt must include {context} placeholder
  • Documents are searched for relevant information
  • Search results replace {context} in the prompt
  • AI responds using your document data
Intro Message: Optional automatic message delivered when users connect

​
Use Cases

  • Customer service representatives with specific guidelines
  • Sales assistants with product knowledge
  • Support agents with troubleshooting procedures
  • Educational tutors with curriculum information

​
Quick Start

  1. Navigate to Resources → Prompts
  2. Click “Add New Prompt”
  3. Enter a name and system prompt
  4. Enable “Require Document” if using RAG (include {context})
  5. Optionally add an intro message
  6. Use the Prompt ID when creating sessions

​
Documents

Documents enable Retrieval-Augmented Generation (RAG), allowing your AI Human to answer questions based on your uploaded data through semantic search. Upload Documents here →

​
Key Features

  • Structured Data Upload: Support for CSV, XLS, and XLSX files
  • Automatic Processing: Documents are split into searchable items
  • Semantic Search: Vector embeddings enable intelligent document retrieval
  • Configurable Search Count: Control how many document items are included

​
How It Works

  1. Upload a structured file (CSV/XLS/XLSX) with headers
  2. System processes the file and creates searchable items
  3. When RAG is enabled in a prompt, user queries trigger document search
  4. Relevant document items are included as context in the LLM prompt
  5. AI generates responses using your document data

​
Quick Start

  1. Navigate to Resources → Documents
  2. Click “Add New Document”
  3. Enter a Title, Description (optional), and “Search Count”
  4. Upload document file.

​
Integration: Using All Three Together

These features work together to create a complete, customized AI Human experience:

​
Example Workflow

  1. Upload Documents: Add your product catalog or knowledge base
  2. Create Prompt: Write a system prompt with RAG enabled ({context} placeholder)
  3. Upload Background: Add a branded background image
  4. Create Session: Combine all three:
    • Use your prompt (with RAG)
    • Select your background image
    • The AI will use your documents to answer questions

​
Summary

  • Background Images: Customize the visual environment of your sessions
  • Prompts: Define AI personality, behavior, and knowledge integration
  • Documents: Provide data for AI to reference through RAG
Together, these features enable you to create highly customized, context-aware AI Human experiences tailored to your specific use cases and brand.