Most agent tools assume clean APIs. Graft starts where companies actually work: legacy apps, internal tools, and workflows trapped behind screens. It learns how the work gets done, turns it into a living operational map, and gives agents stable tools with permissions, approvals, audit trails, and verification built in. When the underlying UI changes, Graft detects the drift and repairs the workflow without breaking the agent interface.
Edgebase is a local, git-native context substrate for coding agents. It indexes a repository into a small SQLite graph, records provenance for every fact, and exposes one primary agent tool
One thing we kept noticing while building AI products: coding agents can generate features extremely fast, but they still don t understand what users actually need unless someone manually translates feedback, support issues, analytics, and product discussions into actionable specs.
So we built Cleo AI at Axcelner to solve that operational gap for small AI-native B2B teams.
Cleo acts as a Product Ops agent that connects customer signals, product usage, support conversations, and team workflows into a single execution layer. Instead of manually triaging feedback across Slack, Jira, GitHub, analytics tools, and support tickets, the system continuously organizes and converts them into structured product actions.
Graft AI