SithraBot/synthetic
🏗️ FrameworkSithraBot
A Kotlin-first LLM toolkit for building AI agents with unified API support and async RAG capabilities.
Synthetic is an innovative Kotlin library engineered to bridge the gap between complex LLM APIs and JVM development. It abstracts the underlying model communication, providing a consistent interface regardless of whether you are using OpenAI, DeepSeek, or custom local models. A core feature of the toolkit is its agent-enabled architecture, which facilitates the creation of autonomous workflows. The framework includes built-in support for asynchronous Retrieval-Augmented Generation (RAG), allowing developers to handle large datasets efficiently without blocking the main execution thread. Its modular design extends to data storage, enabling developers to swap persistence layers seamlessly. By prioritizing a 'zero boilerplate' philosophy, Synthetic reduces the overhead typically associated with setting up chat-based AI agents in Kotlin, making it an excellent choice for developers looking to implement RAG or agentic behaviors in a type-safe, performant environment.
💡Highlights
- ├─Unified API for OpenAI, DeepSeek, HF
- ├─Async RAG with zero boilerplate
- └─Modular data storage architecture
🎯For
- ├─Kotlin Developers
- └─Backend Engineers