The Data Exchange with Ben Lorica The Fenic Approach to Production-Ready Data Processing
Aug 21, 2025
Kostas Paralis, co-founder of Typedef, introduces Fenic, an innovative open-source framework for AI-driven data processing. He unpacks how Fenic treats inference as a key operation, transforming unstructured data management, and optimizing data pipelines. The conversation dives into the evolution of data technologies, the challenges of integrating structured and unstructured data, and the revolutionary potential of large language models. Kostas also discusses the importance of open file formats and practical use cases in cybersecurity and healthcare, enhancing operational efficiency.
AI Snips
Chapters
Transcript
Episode notes
Data Tools Need A First-Principles Rethink
- Existing data platforms were built for different use cases and struggle with modern data needs like ML and embedded analytics.
- Kostas argues we must design tooling from first principles for new workloads and new users.
Make Inference Part Of The Query Engine
- Fennec treats inference as a first-class computation within the query engine rather than a black-box UDF.
- Operators like semantic filtering and joins expose inference so the engine can optimize them.
Store Documents As Typed Markdown Columns
- Store raw converted outputs (like Markdown) in Fennec data frame columns to preserve structure and enable queries.
- Use built-in Markdown column functions to extract headings, sections, and tables of contents across documents.