London-based Conduct, an enterprise AI startup founded by three former Palantir engineers, has raised €10.1M ($12M) in seed funding.
The round was led by Creandum, with participation from Lucid Capital, Booom, and angels from Palantir, Google DeepMind, Workday, and SAP.
Breaking ERP’s bottleneck
The company is building an agentic AI platform that lets enterprises directly communicate with their ERP systems. By making complex codebases and business logic transparent, the platform helps companies modernize faster and cut their reliance on costly external consultants.
The first focus: SAP’s S/4HANA migration, a critical shift that every enterprise must complete before SAP’s 2027 deadline. Companies risk spending millions on failed or delayed projects, with custom code being the biggest sticking point.
Philipp Hoefer, Co-Founder and CPO, said:
“Conduct helps customers evaluate which SAP customisations truly matter, and which can be replaced by standard processes. For every $1 spent on customisation, $2 goes to maintenance. We built Conduct to cut this cost burden while protecting competitive advantage.”
Advertisement
Early traction
Enterprises like Daimler Truck and Rittal are already using Conduct. Its agentic AI learns from each interaction, offering better insights over time while ensuring sensitive data stays private.
Henry Thompson, Co-Founder and CTO, explained:
“Our system is architected so customer data never leaks into outside models. Teams get sharper insights without losing control of their business logic.”
Why now?
Legacy ERP systems have long slowed enterprise innovation. With the 2027 migration deadline looming, CIOs face pressure to balance cost, speed, and reliability. Conduct positions itself as a faster, safer alternative to the traditional consultant-heavy approach.
Peter Specht, General Partner at Creandum, commented:
“Conduct is uniquely positioned to redefine how global leaders manage their most complex systems. Their deep enterprise DNA and relentless focus on customer value stood out immediately.”
