Get started
Documentation for review-first retail teams.
Guides for the product, the review workflow, the data model, the security posture, and the usual places things go sideways.
Start here
These docs explain the product as a working system, not as a pile of screenshots. If you are a retailer, start with the first review. If you are evaluating trust, start with security. If you are trying to understand the model, start with recommendations, scoring, and review.
Start with the right guide
QuickstartWhat happens before a retailer sees the first ranked decision queue.AI and reviewHow recommendations are scored, explained, reviewed, and improved over time.SecurityAuthentication, tenant isolation, rate limits, privacy, monitoring, and recovery.TroubleshootingThe first places to look when access, sync, or the queue does not behave the way it should.
Beyond the basics
What to read next.
Start with the product shape, then move into review habits, workflows, data boundaries, and the limits that keep the product honest.
Weekly reviewHow to make the review short, specific, and useful to a busy retail team.WorkflowsThe practical paths for reorders, stock risk, margin leaks, and catalog cleanup.What data gets usedWhat the product needs to rank decisions, and what it should leave alone.RolesHow owners, buyers, operators, and managers fit into the review loop.Product boundariesWhat these docs explain, and what they do not promise yet.
Learn the concepts
Concept
Meaning here
Where to read
Review-first
Recommendations are surfaced for human approval, not automatically executed.
Human approval
Decision layer
The product sits above existing retail systems and helps decide what deserves attention.
What it is
Guided connectors
Connections are checked for scopes, fields, quality, and operational fit before routine use.
Connectors
Learning loop
Approvals, skips, edits, and corrections improve future ranking.
Learning loop