Fenic just dropped on GitHub, a PySpark-inspired, open-source engine where LLM inference is just another column op (map, classify, extract, semantic.join).
Why it’s interesting:
Merges data prep + batch inference + eval into one API
Great for dataset augmentation, schema extraction, and RAG joins
Still early: Rust toolchain, thin docs, light ecosystem
We rated it 4/5 in our latest AI/ML Tool Review.
Full breakdown here → https://neurlcreators.substack.com/p/fenic-dataframes-for-an-llm-world
Would you adopt something like Fenic for your LLM workflows, or stick with Python scripts and SDK calls? Drop your take in the comments. We would love to hear how you’re handling data + inference today.