Getting Started with Kinexa AI
Getting Started with Kinexa AI
What is Kinexa AI Gateway?
Kinexa AI Gateway is the unified AI layer that powers intelligent features across all Kinexa modules — from Liqaa's AI chatbot to Wawas's AI insights. Instead of integrating with each AI provider separately, Kinexa routes all AI requests through a single gateway that handles provider selection, fallback, rate limiting, and cost optimization.
Available AI Providers
Kinexa supports multiple AI providers through the gateway:
| Provider | Models | Type |
|---|---|---|
| Anthropic Claude | Haiku, Sonnet, Opus | Cloud API |
| OpenAI | GPT-4o, GPT-4o-mini | Cloud API |
| Ollama | Qwen 2.5, Llama 3.2, Gemma 2, Phi-3, Mistral | Self-hosted |
AI Assignment per Pricing Tier
Each Kinexa pricing tier comes with a default AI provider:
| Tier | Default AI Model | Provider |
|---|---|---|
| Free / Starter | Qwen 2.5 7B | Ollama (self-hosted) |
| Pro | Claude Haiku | Anthropic |
| Business | Claude Sonnet | Anthropic |
| Enterprise | Claude Opus | Anthropic |
Higher tiers get access to more capable models. Enterprise customers can also request custom model configurations.
How to Configure AI in Liqaa
- Login to your Liqaa member dashboard
- Go to Settings > AI Chatbot
- You will see your current AI model based on your tier
- Configure AI behavior:
- AI Name — The name your chatbot uses
- AI Personality — Tone and style (professional, friendly, casual)
- Custom Instructions — Specific rules for the AI (e.g., "Always recommend our Pro plan")
- Handoff Rules — When to transfer to a human agent
- Click Save
Tips
- Free/Starter users get fast, capable responses from Ollama models at zero per-request cost
- Upgrade to Pro or Business for more sophisticated AI responses
- Enterprise customers can use dedicated Ollama instances with larger models