
Typedef is an AI-native data infrastructure company that provides a data engine designed to power production-ready AI workloads, particularly LLMs. It aims to bridge the gap between traditional data stacks and the demands of modern AI, focusing on semantic processing, operationalizing AI workflows, and bringing structure to unstructured data. Typedef's engine is built with a serverless, inference-first approach, using Rust-powered compute for efficiency and Python simplicity. It allows developers to transform unstructured and structured data using familiar DataFrame operations, similar to PySpark or Pandas, enabling tasks like semantic joins and schema-driven extraction. The company emphasizes moving from prototype to production with confidence, backed by experienced infra builders and AI/data practitioners from companies like Snowflake and Salesforce. They offer an open-source library called Fenic for local development.

Typedef is an AI-native data infrastructure company that provides a data engine designed to power production-ready AI workloads, particularly LLMs. It aims to bridge the gap between traditional data stacks and the demands of modern AI, focusing on semantic processing, operationalizing AI workflows, and bringing structure to unstructured data. Typedef's engine is built with a serverless, inference-first approach, using Rust-powered compute for efficiency and Python simplicity. It allows developers to transform unstructured and structured data using familiar DataFrame operations, similar to PySpark or Pandas, enabling tasks like semantic joins and schema-driven extraction. The company emphasizes moving from prototype to production with confidence, backed by experienced infra builders and AI/data practitioners from companies like Snowflake and Salesforce. They offer an open-source library called Fenic for local development.
Product: Serverless, inference-first AI data engine for production-ready LLM/AI workloads
Founded: 2023
Funding: $5.5M seed (Jun 18, 2025) led by Pear VC
Founders: Yoni Michael; Kostas Pardalis
Employees: 8
Productionizing AI workloads that require semantic processing of unstructured data and integrated LLM inference within analytical workflows.
2023
AI data infrastructure
5500000.00
Participating investors listed include Monochrome Capital, Verissimo Ventures, and Tokyo Black; announcement also referenced additional angel participants.
“Seed round led by Pear VC with participation from Monochrome Capital, Verissimo Ventures, and Tokyo Black.”