What is PebbleAI?

What is PebbleAI?

PebbleAI is the AI consolidator for your team or business — an enterprise AI platform that replaces the ten scattered AI tools your staff are currently using with one sanctioned place to chat, build, and observe. Whether you’re looking to have AI-powered conversations, build custom AI workflows, or deploy intelligent agents, PebbleAI provides all the tools you need behind one boundary you control.

Why choose PebbleAI?

Your data stays inside your organisation. Model calls route through your own credentials — typically AWS Bedrock in your chosen region (e.g. ap-southeast-2 for Australia) — so conversations, documents, and AI-generated content never leave the boundary your organisation controls. Data residency, contractual coverage, and compliance regimes like IRAP PROTECTED are respected automatically, without anyone on your team having to think about it.

One platform, one audit trail. Instead of ten AI tools with ten separate logins, ten separate bills, and ten separate security reviews, PebbleAI is the single sanctioned place. Admins see every interaction. Finance sees every dollar. Security sees every boundary.

Consolidation without compromise. Users still get the best model for each job — Claude Opus for reasoning, Claude Haiku for speed, the right model automatically via PebbleRouter — because PebbleAI connects to every major provider. You don’t give anything up by centralising.

Skills that actually execute code. PebbleAI ships with Skills — reusable prompt templates that are byte-compatible with the Claude Code skills spec and can include Python code that runs in a platform-hosted, gVisor-isolated code execution sandbox. A skill can produce a polished branded PDF, an Excel spreadsheet, a chart, a data analysis — and the generated file lands in your user file store and comes back as a download link in the chat. Claude, OpenAI, and Copilot don’t do this inside enterprise chat.

Open, not locked in. Built on Flowise and Langfuse (both leading open-source projects), PebbleAI’s flows and traces live in formats you can export. No vendor can hold your work hostage.

Key Capabilities

PebbleChat - Your AI Agent

More than a chatbot, PebbleChat is an AI agent you work with to get things done. Connect to leading AI models from OpenAI, Anthropic, Google, and AWS Bedrock. Use @mentions to reference your workflows, attach files for analysis, speak with voice input, connect external tools via MCP, run Skills that execute Python code in a sandbox, and let PebbleChat auto-discover the right resources for your task. Export responses as PDF, Word, or email, and let background processing handle long-running tasks.

PebbleFlows - Visual AI Builder

Create sophisticated AI workflows without writing code. Our drag-and-drop interface lets you:

  • Build conversational chatbots and AI assistants
  • Create document processing pipelines
  • Design multi-agent systems
  • Connect to your existing tools and databases
  • Share workflows across workspaces with asset discovery

PebbleAgents - Enterprise Agents (Coming Soon)

Autonomous AI agents that can plan, collaborate in teams, manage tasks, and execute complex workflows with minimal supervision. PebbleAgents will bring enterprise-grade autonomous capabilities to your organisation.

PebbleObserve

Monitor and optimise your AI applications with comprehensive observability:

  • Real-time trace analysis
  • Performance metrics and cost tracking
  • Quality monitoring and debugging
  • Usage analytics and insights

Team Collaboration

Work together on AI projects with built-in collaboration features:

  • Shared workspaces
  • Team chat histories
  • Collaborative workflow editing
  • Role-based access control

Enterprise Ready

Designed for organizations of all sizes with:

  • Multi-tenant architecture
  • Secure credential management
  • Usage monitoring and analytics
  • Custom branding options
  • SSO integration

Who Uses PebbleAI?

Business Teams

  • Customer Support: Build AI assistants to handle common inquiries
  • Marketing: Generate content and analyze campaign performance
  • Sales: Create lead qualification bots and proposal generators
  • HR: Automate FAQ responses and initial candidate screening

Technical Teams

  • Developers: Rapidly prototype AI features
  • Data Scientists: Deploy models without infrastructure concerns
  • IT Teams: Manage AI access and monitor usage across the organization

Individual Users

  • Knowledge Workers: Get AI assistance for daily tasks
  • Content Creators: Generate and refine content
  • Analysts: Process and analyze data more efficiently

How It Works

  1. Choose Your Tool

    • Use PebbleChat for immediate AI assistance
    • Open the Visual Builder to create custom workflows you can share
  2. Select AI Models

    • Pick from pre-configured models
    • Or bring your own API keys or LLMs
  3. Build or Chat

    • Start conversations in PebbleChat
    • Design workflows with drag-and-drop components
  4. Deploy and Share

    • Share chatflows with your team
    • Embed AI capabilities in your applications
    • Monitor usage and performance

Next Steps