Devgraph.ai
Devgraph connects engineering tools into an ontology: a model of the relationships between code, infrastructure, systems and teams. Its intended audience is an engineering organization whose service knowledge lives across repositories, tickets, chat and operational tools.
Connect the engineering stack
The publisher lists GitHub, GitLab, Jira, Vercel, Kubernetes, Argo, FOSSA, Grafana, Slack and PagerDuty as sources. A flexible API supports custom connections. Devgraph builds its ontology from the connected stack, then lets teams query across those sources from one place.
That source coverage matters when evaluating the product. A service's repository may explain its implementation while a ticket or discussion explains why a change happened. Devgraph's stated approach connects those records through their relationships, so the useful question is whether it represents the systems and team ownership your organization actually uses.
Investigate dependencies and decisions
Devgraph is designed to show services that depend on a proposed change and the teams that may need an alert. It also surfaces decisions from Slack and links discussions to code and tickets. The publisher describes onboarding questions about ownership, dependencies and deployment as another use case.
These are documented aims, not proof that the platform will predict every incident. Try questions with known answers when assessing the ontology: whether it identifies the owner of a service, connects a decision to the right ticket, and reflects the dependencies relevant to a particular change.
Choose a model and confirm deployment requirements
Devgraph supports hosted models and self-hosted or air-gapped deployments, according to the publisher. Its site names OpenAI, Anthropic, xAI and Ollama, and describes switching providers while keeping the workflow.
The captured official page does not give prices, permission rules, setup requirements or refresh intervals. Before adopting it, confirm how source access and ownership are represented, which systems can connect in your deployment, and how quickly changes reach the ontology. Those details determine whether cross-system answers fit your team's operational needs.