Conduct raises $60M to build agentic enterprise OS

Conduct, a London-based startup founded in 2024 by former Palantir engineers Philipp Hoefer, Henry Thompson and JP Haas, has raised $60M in a Series A round co-led by Index Ventures and Iconiq, Sifted reports. SAP provided a strategic investment and is embedding Conduct in its products, while Creandum, Lucid Capital and Bloom also participated. The raise follows an $11M seed round closed less than nine months earlier, per Sifted; Tech.eu reports total funding of around $72M. Per Sifted and Tech.eu, Conduct builds an agentic AI operating system that analyses code and configurations across enterprise systems to make business logic legible and actionable. Reported customers include Daimler Truck, Heidelberg Materials and DHL. CEO JP Haas told Tech.eu: "The same opacity that slows people down stops agents entirely, because an agent can only act on a system it understands. Conduct makes those systems legible and operable."
What happened
Conduct has raised $60M in a Series A round co-led by Index Ventures and Iconiq, Sifted reports. Creandum, Lucid Capital and Bloom participated, and SAP provided a strategic investment while embedding Conduct in its products, according to Sifted. The raise follows an $11M seed round closed less than nine months earlier, per Sifted; Tech.eu puts the total at around $72M. The company was founded in 2024 by three former Palantir engineers: Philipp Hoefer, Henry Thompson and JP Haas.
Technical details
According to Sifted and Tech.eu, Conduct analyses code and configurations across enterprise systems -- particularly SAP deployments -- to synthesise and surface business logic that has accumulated through decades of customisation. The platform then makes those systems legible and actionable for AI agents, human teams and external consultants. CEO JP Haas told Tech.eu: "Every major enterprise is being asked where its AI results are. The honest answer, in most organisations, is that the systems AI needs to work on today cannot be fully comprehended by humans. Decades of customisation have made them opaque, even to the people running them. The same opacity that slows people down stops agents entirely, because an agent can only act on a system it understands. Conduct makes those systems legible and operable. That is the foundation everything else depends on." Sifted lists Daimler Truck, Heidelberg Materials and DHL as reported customers; the team is around 35-38 people in London with a stated target of 100 by year end and a new New York office opening.
Industry context
Agentic systems that surface and act on enterprise business logic are an emerging category combining program analysis, knowledge extraction and automation orchestration. Investors and platform vendors placing bets in this space generally seek two things: technical capability to map bespoke customisations across SAP, Oracle and similar platforms, and distribution through partnerships with incumbent vendors. SAP's strategic investment and product embedding, as reported by Sifted, aligns with a pattern where platform vendors accelerate adoption by pre-integrating AI-driven automation tools.
What to watch
Monitor integration depth and observability around automated change actions, because AI-driven modifications to mission-critical ERP and supply-chain codebases raise operational risk and auditability requirements. Watch for published customer outcomes from Daimler Truck and other major adopters, and for independent evaluations of how well agentic approaches preserve business intent while making safe changes.
Scoring Rationale
A $60M Series A for an enterprise agentic OS that targets ERP-system opacity is notable: the SAP strategic investment and product embedding provide meaningful distribution, and the technical thesis -- making legacy ERP customisations machine-readable -- addresses a concrete bottleneck to enterprise AI adoption. Notable but not transformative; scores in the mid-range for a well-backed European enterprise AI funding round.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
