We are building the universal runtime for AI agents.
As foundation models cross the threshold of reliable autonomous reasoning, an entire layer of human cognitive labor becomes economically obsolete. Mandarum is the platform that operates that layer — safely, observably, and at enterprise scale.
Make autonomous work trustworthy.
We believe the agent era will be defined by reliability and governance — not raw model capability. Our job is to give every enterprise the runtime, identity, and policy infrastructure to run agents like first-class citizens of the org.
What we hold true.
Reliability is a feature.
Demos are easy; production is not. We invest disproportionately in evals, traces, and recovery.
Policy before action.
An agent without guardrails is a liability. Approval is a first-class primitive.
Open by default.
Bring any model, any framework, any cloud. Lock-in lives in workflows, not protocols.
Design is the moat.
In an era of commodity intelligence, craft is the differentiator.
Ship weekly.
Velocity compounds. Enterprises that adopt Mandarum get a new product every Friday.
Customer outcomes > ours.
We measure ourselves in dollars and hours saved for the people who pay us.
Operators and patient capital.
Builders from the systems that move the world.
Maya Okafor
CEO · ex-Stripe, Anthropic
Jonas Vetter
CTO · ex-Temporal, Databricks
Priya Subramanian
Head of Design · ex-Linear
Daniel Park
Head of AI · ex-OpenAI
Renata Lima
Head of Security · ex-Vanta
Oliver Reed
Head of GTM · ex-Snowflake
Hannah Schulze
Founding PM · ex-Notion
Yusuf Ahmadi
Founding FDE · ex-Palantir