Deriv has released Deriv One, a self-hosted enterprise AI platform, as open source under the MIT licence on 29 October 2026. The platform enables any employee to access AI capabilities aligned with their role, with built-in governance controls. No vendor account or payment is required to deploy the system.
- Skills developed: 221 skills from 54 employees across the company
- Deployment target: Automating 75% of manual workflows by the end of 2026
- Core features: Workspace with split panes, role-based skills catalogue, agent registration, and Lab for tool deployment
- Technical requirements: Three commands to deploy; compatible with OpenAI-compatible gateways, LM Studio, or Ollama
- Company workforce: Over 900 employees across 13 countries
The platform originated from CEO Rakshit Choudhary‘s request for a skills registry. As AI agents proliferated across Deriv, individual teams developed their own skills without clear visibility into who could access or run which tools. Deriv One centralises this functionality while maintaining role-based access controls.
AI shouldn’t be limited to the few people who can set up their own agent. With Deriv One, anyone at the company can use the skills their role allows, and the guardrails are already built in.
– Rotimi Akinyele, Chief Technology Officer, Deriv
The open-source release includes sandboxing, Docker, and Kubernetes manifests with guides for GKE and EKS deployment. Deriv‘s production cloud plugins and internal integrations remain proprietary; adopters must supply their own model gateway, identity provider, and cloud infrastructure. The platform abstracts away the foundational layer, identity management, secrets handling, skills review, and sandboxing, so other organisations don’t have to rebuild these components. Deriv, which has served over 3 million traders globally for 26 years, continues to invest in AI literacy at all organisational levels as part of its broader AI-first transformation.
Rotimi Akinyele, Chief Technology Officer at Deriv. Source: LinkedIn
