Shadai Client Documentation#
Welcome to the complete documentation for the Shadai Python client! This documentation will help you build powerful AI applications with document understanding, intelligent agents, and RAG technology.š Documentation Structure#
Main documentation hub with complete navigationš Quick Links#
New to Shadai?#
Start here to get up and running:Building Applications?#
Understanding Shadai?#
Real-World Use Cases?#
API Reference?#
Complete technical reference:Code Examples?#
Advanced Topics?#
Take it to the next level:šÆ Common Tasks#
I want to...#
Choose specific AI models ⨠NEWAnalyze multiple documentsš Learning Paths#
Path 1: Build a Q&A System (30 minutes)#
Path 2: Master Core Features (1 hour)#
Path 3: Build Custom Agents (2 hours)#
Path 4: Production Deployment (1.5 hours)#
š Search by Topic#
RAG & Document Processing#
Memory & Conversations#
Integration & Production#
š” Tips#
Start Simple: Begin with basic queries, add complexity as needed
Use Memory: Enable memory by default for natural conversations
Stream Responses: Always stream for better user experience
Handle Errors: Use try/except for robust applications
Reuse Sessions: Keep sessions alive for better performance
š Need Help?#
š What's New in v0.1.30#
šÆ Model Selection - Choose from 31 LLM models and 5 embedding models across OpenAI, Azure, Anthropic (Claude), and Google (Gemini)
ā” Automatic Error Handling - Clean, user-friendly error messages without verbose tracebacks
šØ System Prompts - Customize your session's behavior with custom system prompts
š§ Provider Flexibility - Mix and match providers (e.g., Google LLM + OpenAI embeddings)
Previous Release (v0.1.29)#
š§ Memory Enabled by Default - All tools now use conversation memory automatically
š¬ Chat History Management - New methods to retrieve and clear session history
š Complete Documentation - This comprehensive documentation site!
š License#
MIT License - see LICENSE file for details.
Modified atĀ 2025-10-21 05:36:37