
"Ona's platform introduces high-autonomy agents that operate in fully sandboxed environments, each with OS-level isolation to avoid cross-environment interference. These Ona Agents outperform previous solutions based on their own experience. Inside its own engineering teams, they co-authored 60% of merged pull requests and contributed 72% of merged code last week. Meanwhile, according to the company, customers report a 4x increase in development throughput. The company has even extended usability across devices. Ona works via browser-based VS Code, on smartphones, or in desktop IDEs, making development seamless across contexts."
"Ona offers an ecosystem comprised of three core components: (1) Ona Environments, which are API-first, sandboxed developer environments preconfigured with dependencies and integrations and defined via devcontainer.json and automations.yml; (2) Ona Agents, AI-powered collaborators supporting multiple workflows and device types, with slash commands to automate common engineering tasks; and (3) Ona Guardrails, offering enterprise-grade security including audit trails, RBAC, SSO/OIDC, command-level controls, and deployment in user-defined VPCs."
Ona repositions cloud development toward AI-driven software engineering by providing autonomous agents, sandboxed developer environments, and security guardrails. High-autonomy agents run with OS-level isolation to prevent cross-environment interference. Internal adoption resulted in agents co-authoring 60% of merged pull requests and contributing 72% of merged code, while customers report up to 4x increased development throughput. The platform supports browser-based VS Code, mobile, and desktop IDEs. The ecosystem comprises Ona Environments (API-first, preconfigured devcontainers), Ona Agents (AI collaborators with slash commands), and Ona Guardrails (audit trails, RBAC, SSO/OIDC, command controls, VPC deployment).
#ai-driven-software-engineering #autonomous-agents #sandboxed-developer-environments #enterprise-security
Read at InfoQ
Unable to calculate read time
Collection
[
|
...
]