← All organizations

Enterprise AI orchestration and development

deepset

deepset builds software for developers and enterprise teams to create AI agents, document search, and retrieval-augmented generation applications. Its Haystack open-source Python framework connects models, retrieval, tools, and memory into modular pipelines. Haystack Enterprise Platform adds visual development, testing, observability, access controls, and deployment to managed cloud or self-hosted infrastructure. Enterprise Starter pairs the framework with engineering support and deployment guidance. In 2025, deepset renamed its commercial deepset AI platform to Haystack Enterprise Platform, bringing its offerings under the Haystack product family.

Founded in 2018 by Milos Rusic, Malte Pietsch, and Timo Möller, deepset is led by Rusic as CEO and Pietsch as CTO. Its early work building custom language AI systems informed Haystack, developed in 2020. The framework’s lasting contribution is its composable architecture: developers can inspect, test, and replace individual retrieval, routing, and generation components, adapting models and data stores without rebuilding the whole application.

As of 2026, deepset reported that the European Commission’s AI@EC platform supported 68 builders across 29 departments, with 316 pipelines in staging or production. The company raised $30 million in 2023 in a funding round led by Balderton Capital.

www.deepset.ai

1 talk

Newest first

1 speaker at AIE

Affiliations reflect their AIE appearances, not necessarily current employment.

Messages from the stage

Opponent difficulty and training stability

A tic-tac-toe environment with minimax-based opponents provides a concrete setting for examining opponent difficulty, rollout inspection, sampling temperature, and instability related to batch size.

Affiliations reflect each recorded session, not necessarily current employment.

Company sources · checked 2026-08-27