IBM/Agentics: Agentics 0.3.0: Semantic Operators & Enhanced Documentation

Agentics 0.3.0: Semantic Operators & Enhanced Documentation Welcome to Agentics 0.3.0! This release introduces powerful new semantic operators and significantly enhances the framework's documentation and developer experience. About Agentics Agentics is a Python framework for structured, scalable, and semantically grounded agentic computation. Unlike traditional agent frameworks that pass untyped text through pipelines, Agentics flips the paradigm: types are the interface. Build AI-powered pipelines as typed data transformations—combining Pydantic schemas, LLM-powered transduction, and async execution. Core Philosophy: "Transduction is all you need" — workflows as composable, type-safe transformations between structured states. 🎯 What's New in 0.3.0 ✨ Semantic Operators Module Introducing a new semantic operators module with powerful operations for semantically-grounded transformations: sem_map — Apply semantic transformations across collections sem_filter — Filter data based on semantic criteria using LLMs sem_agg — Aggregate and summarize data semantically These operators enable richer, more expressive workflows while maintaining type safety and semantic grounding. 📚 Improved Documentation Comprehensive documentation updates across tutorials and guides New Discovery Demo showcasing real-world use cases Enhanced README with clearer examples and concepts Better getting-started experience for new users 🔧 Integration Fixes Resolved Langflow integration issues Code quality improvements (whitespace normalization, file endings) 🚀 Key Features (Full Framework) Typed agentic computation — Define workflows over Pydantic schemas Logical transduction (<<) — Transform data between types using LLMs Async mapping & reduction — Scale with amap and areduce Batch execution & retry — Built-in batching and graceful fallbacks Tool support (MCP) — Integrate external tools seamlessly Semantic operators — NEW: semantically-grounded transformations 📦 Getting Started ```bash uv pip install agentics-py ``` Set up your `.env` with your LLM provider credentials, then try: ```python from pydantic import BaseModel, Field from agentics.core.transducible_functions import Transduce, transducible class Movie(BaseModel): movie_name: str description: str year: int class Genre(BaseModel): genre: str = Field(description="e.g., comedy, drama, action") @transducible async def classify_genre(state: Movie) -> Genre: """Classify the genre of a movie.""" return Transduce(state) Use it genre = await classify_genre(Movie(...)) ``` 📖 Learn More Tutorials: Comprehensive Jupyter notebooks in `tutorials/` covering core concepts, semantic operators, and advanced patterns Documentation: Full reference at `docs/index.md` Examples: Runnable examples in `examples/` folder Paper: "Transduction is All You Need for Structured Data Workflows" — arXiv:2508.15610 🙏 Thanks Special thanks to all contributors and the community for feedback, issues, and pull requests that make Agentics better every release. Framework: Python 3.11+ License: Apache 2.0 Repository: https://github.com/IBM/Agentics