ai-worker, Part 1: From Issue to Reviewed Merge
The current ai-worker workflow: three bot roles, approved sequential milestones, evidence-based review and repair, CI-gated delivery, and a web control plane.
Read article →
AmitableConnected data and workflowsThe Amitable notebook
Amitable blog
Notes on Amitable, its architecture, and connected business data.
ai-worker, Part 2: Architecture and Execution Boundaries
Inside the current ai-worker: a shared scheduler, ordered provider adapters, a protected web control plane, isolated QEMU execution, snapshot-bound evidence, and delivery cleanup.
Read article →Automate routine steps
Set up actions around new records, status changes, and dates in a workflow.
Read article →Build collections around your work
Shape collections and fields around customers, projects, orders, and other records your team manages.
Read article →Connect related records
Link records across collections so customers, orders, products, and suppliers stay connected.
Read article →Create filters and chart previews with AI
Describe the records you want to find or the values you want to compare in a collection.
Read article →Import existing data with a CSV
Start a collection from a CSV file and review its rows and columns before bringing records into Amitable.
Read article →Work together in real time
See colleagues' changes while working in the same collection and keep shared records current.
Read article →Building Amitable: architecture and engineering principles
The engineering ideas behind Amitable: explicit boundaries, typed data, repository APIs, and observable workflows.
Read article →What Amitable is and the problem it solves
Why Amitable puts connected business data, workflows, and AI assistance in one workspace.
Read article →Workspaces, collections, relationships, and connected business data
How Amitable turns separate business objects into a navigable system of connected records.
Read article →