DOCSEARCHbyAlgolia
Submit Docs

Sairis Documentation

website·Indexed 30 days ago

https://help.sairis.ai

User and administrator guides for the Sairis platform, featuring tools for AI Builder workflow creation, Knowledge Base document processing, and natural language data analytics. Includes detailed instructions on user and role management, system administration, and AI foundation model comparisons.

Tokens
72.5K
Snippets
124
Records
577
Agent score
49%

What's inside Sairis

  1. Overview of LaunchPad features

    LaunchPad is a customizable home screen for the AI Assistant designed to provide quick access to frequently used content. It allows administrators to create content sections, filter content by type, topic, or tags, search across all available content, and prioritize information by reordering sections.

  2. Overview of Topic Management in Sairis AI

    Topics are centrally managed primary categories used to provide a structured organizational system for content across the platform. Unlike tags, each content item is assigned a single primary topic. Topics are used to facilitate content discovery, organize AI Blueprints and Knowledge Bases, and generate high-level analytics and reporting on user interactions.

  3. Overview of Prompt Engineering in Sairis

    Prompt engineering in Sairis is the practice of designing and refining prompts to obtain precise, relevant, and structured AI responses. It involves controlling AI behavior, managing context, and leveraging knowledge retrieval to optimize AI for complex business workflows.

  4. Overview of Knowledge Base Tool capabilities

    The Knowledge Base Tool is an intelligent assistant for processing documents, PDFs, and text content. It allows users to discover, search, summarize, compare, and explore document collections using natural language queries. Key capabilities include listing available documents, performing semantic searches, summarizing full documents, comparing multiple documents, generating textual diffs, and caching documents to prevent redundant loading.

  5. Overview of AI Builder capabilities

    AI Builder is a platform designed to transform users into AI experts by removing the steep learning curve of manual prompt engineering. It enables the creation of professional-grade AI assistants in hours rather than months by providing guided intelligence, a library of proven templates, and continuous AI-powered optimization. Key benefits include a 95%+ consistency rate across scenarios and significant improvements in accuracy, relevance, and task success rates compared to manually crafted prompts.

  6. Overview of Sairis End User capabilities

    As an End User of Sairis, you are the primary consumer of the AI assistant platform. Your role involves using the system to ask questions, generate ideas, summarize documents, and interact with assigned content. You have the ability to control AI instructions, select different AI models, and utilize tools such as web and news searches.

  7. Configure basic Knowledge Base information

    In the KB Settings tab, define the fundamental properties of the knowledge base:

    • Name: A required descriptive title used in selection menus and search results.
    • Initials: 1-2 characters for the avatar (auto-generated from the name if left blank).
    • Description: Details about the purpose and content of the KB to help users.
    • Topic: A category selected from predefined organizational topics for filtering.
    • Tags: Comma-separated keywords to improve discoverability.

  8. Implement AI style using an iterative refinement process

    Apply communication styles in three progressive phases to optimize user interaction:

    • Phase 1: Basic Implementation: Establish fundamental tone and vocabulary, test core use cases, and gather initial feedback.
    • Phase 2: Style Refinement: Adjust based on real user interactions, add guidelines for edge cases, and expand vocabulary variety.
    • Phase 3: Advanced Optimization: Implement context-aware adaptations, fine-tune for specific user segments, and add cultural/regional considerations.

  9. Understand User Roles in Sairis AI

    User Roles are used to group users for reporting purposes and to associate specific features and settings with groups of users. Modifying a Role automatically updates all users associated with that role. For users integrated via HRIS or HCM, default settings can be established for Roles so that new users inherit those settings upon creation.

  10. Avoid common AI role definition mistakes

    To ensure high-quality AI interactions, avoid vague or overly broad definitions. Be specific about the AI's purpose and the platform it supports.
    # Avoid Vague Definitions
Bad: "You are a helpful assistant who answers questions."
Better: "You are a customer support specialist for a SaaS platform who helps users troubleshoot technical issues, explains features clearly, and guides users toward successful outcomes."

# Avoid Overly Broad Expertise
Bad: "You are an expert in everything and can help with any topic."

  11. Understand the Sairis Business Admin role and permissions

    The Business Admin role is designed for managing departmental or functional area settings.

    Key Responsibilities:

    • Creating and managing department-specific knowledge bases, blueprints, and content.
    • Establishing and maintaining department Smart Groups.
    • Creating and managing users within their specific department.
    • Monitoring usage of departmental resources.

    Permissions and Limitations:

    • Scope: Can only manage content assigned to their specific department.
    • User Management: Can only handle users within Smart Groups or roles for which they have been granted edit capabilities.
    • Restrictions: Prohibited from creating new roles or modifying global system settings.

  12. Understand the types of AI instructions in AI Builder

    AI instructions are detailed guidelines that define how an AI system processes input, prioritizes information, structures responses, handles conditional logic, and respects safety boundaries. They are categorized into five main types:

    • Processing Instructions: Step-by-step procedures for handling information.
    • Decision Instructions: Logic for choosing between different approaches.
    • Quality Instructions: Standards for output quality and accuracy.
    • Safety Instructions: Boundaries and restrictions to maintain appropriate behavior.
    • Format Instructions: Specifications for how responses should be structured.