Opengeni
Opengeni is an open-source, self-hostable runtime and managed platform for embedding AI agents that do real work in products. It supplies durable sessions, execution environments, tools and integrations, approvals, memory, observability and multi-tenant infrastructure so agents can continue working safely across failures and over long periods.
At a Glance
- Software and SaaS product teams
- Enterprise engineering and platform teams
- Developers building AI-enabled products
- Organizations needing self-hosted or private agent infrastructure
- +1 more
AI Tools by Opengeni
(1)
Opengeni
Open Source AI Agent Runtime
Discussions
No discussions yet
Be the first to start a discussion about Opengeni
Latest News
Opengeni launches on Product Hunt as infrastructure for shipping AI agents to production
Opengeni Terms of Service updated; terms identify Cloudgeni AS as the service provider and describe the hosted platform and credit billing
opengeni-agent v0.1.49 published as a cross-platform signed release
Signed native artifact runtime canary releases published for the Opengeni repository
Products & Services
Fully hosted service for running agent sessions with managed infrastructure and updates, supported models, sandboxed execution, audit and usage controls, and team workspaces.
Apache-2.0 software that organizations can run locally or self-host on their own infrastructure, including the API, web app, workers, deployment artifacts and reference infrastructure configuration.
TypeScript SDK, server-side chat/session proxy and React components for embedding agent conversations and controlling sessions, workspaces, tools, files and event streams in another product.
Plugin/skills for Claude Code, Codex and Cursor that helps a coding agent integrate Opengeni into a product.
Market Position
Opengeni positions itself as the infrastructure/runtime layer around an AI agent rather than another model or standalone agent: durable orchestration, sandboxed compute, approvals, governance, integrations, memory and observability in one open-source/self-hostable stack. Its main alternatives are build-your-own agent runtimes and orchestration platforms such as LangGraph/LangSmith, Temporal-based agent systems, Microsoft Azure AI Agent Service, Google Vertex AI Agent Builder and AWS Bedrock Agents; Opengeni differentiates with Apache-2.0 self-hosting, an embedded product UI/SDK, connected-machine execution and a managed cloud priced at model cost plus 5%.
Leadership
Founders
Davlet Dzhakishev
CEO/CTO and co-founder of Cloudgeni, the team behind Opengeni; former Microsoft engineer and startup CTO.
Iuliia Petryshyn Thuen
COO and co-founder of Cloudgeni; former McKinsey and Accenture consultant and former COO at procurement scale-up Ignite.
Executive Team
Davlet Dzhakishev
CEO/CTO & Co-Founder, Cloudgeni (Opengeni creator)
Former Microsoft engineer and startup CTO; built Cloudgeni after experiencing infrastructure toil.
Iuliia Petryshyn Thuen
COO & Co-Founder, Cloudgeni (Opengeni creator)
Former McKinsey and Accenture consultant and former COO at Ignite.
Founding Story
Opengeni grew out of two years of Cloudgeni running AI agents against production cloud infrastructure. The team found that making an agent work once was only the beginning: durable execution, safe sandboxes, credentials, human approvals, integrations, memory, auditability and cost visibility were the recurring infrastructure work. They extracted that infrastructure and open-sourced it so developers could ship production agents rather than rebuild the surrounding runtime.
Business Model
Revenue Model
Self-hosted software is free under Apache 2.0; the hosted Opengeni Cloud charges provider model cost plus 5%, with no platform fee, minimum commitment or per-seat fee. Enterprise deployments are custom-scoped and may include commercial support and deployment/integration help.
Pricing Tiers
Self-host on your own infrastructure; customers pay their own compute, storage and model-provider costs.
Managed infrastructure, supported models, sandboxed execution, audit and usage controls; no seat fees, platform fee or minimum commitment.
Private infrastructure, deployment/integration assistance, security/access requirements, SSO and commercial support options.
Target Markets
- Software and SaaS product teams
- Enterprise engineering and platform teams
- Developers building AI-enabled products
- Organizations needing self-hosted or private agent infrastructure
- Teams operating workflows with sensitive systems, credentials or compliance requirements
- Embedding agents inside SaaS products and internal applications
- Long-running background work triggered by schedules or inbound webhooks
- Customer support and ticket-resolution agents
- Agents that call a product's APIs and act as the signed-in user
- Software engineering tasks such as running tests, changing code and opening pull requests
- Operational workflows requiring sandboxed execution, approvals and auditability
- Hydro
- Havila
- Posten