Typedef offers a purpose-built, inference-first AI data infrastructure that manages LLM-powered pipelines, unstructured data processing, and batch AI workloads with minimal operational overhead. The serverless platform exposes a developer-friendly, open-source client library that connects to data sources and handles token limits, context windows, chunking, and other inference complexities through familiar APIs and relational models. Typedef positions itself to move enterprise AI projects from prototype to production faster by enabling deterministic, scalable workflows on top of non-deterministic LLMs and supporting agentic and OLAP-enhanced workloads. The company emphasizes rapid, iterative prompt and pipeline experimentation to identify production-ready workloads and then scale them reliably. Customer testimony from Matic highlights faster deployment of semantic extraction pipelines, cost reductions, and lower E&O risk. Typedef emerged from stealth concurrent with a $5.5M seed raise and is led by co-founders Kostas Pardalis and Yoni Michael, the latter having sold his prior company Coolan to Salesforce in 2016.

Total raised

$6M

Funding rounds

1

Latest round

Seed

Latest activity

Jun 2025