One screen for the people who own the return number.
A live operations view of margin-at-risk, recovery rate, fraud signals, queue depth, and weekly trend. Built so a returns lead can answer 'how are we doing?' in under five seconds.
- Margin-at-risk and recovery rate up top
- Live queue depth by reason and brand
- Drill into any tile to a filtered queue
The cases that matter, sorted by what they actually cost.
Multi-item returns prioritized by margin-at-risk, not arrival time. Bulk approvals on safe cases, deep review on the rest. Saved views per reviewer.
- Multi-item return support throughout
- Saved views, filters, and bulk actions
- Per-item reasons, conditions, and routing
Patterns surfaced as signals — never as labels.
Velocity, value mismatch, geo and address mismatch, photo mismatch, drop-window risk. Each signal is a recorded reason a reviewer can confirm or override.
- Per-case scoring with the reasoning visible
- Reviewer override with audit trail
- No customer-facing labels, ever
Photo-grounded condition decisions, in the warehouse flow.
Inspectors capture evidence, mark condition grade, and the AI suggests a grade with confidence and a defect tag list. The reviewer keeps the final call.
- Grade A through F with consistent definitions
- AI grade as a suggestion, not a decision
- Photo evidence pinned to the case
Refund, store credit, or exchange — recommended, not enforced.
For each return the engine recommends a method with a reason: customer LTV, item resellability, declared reason, brand policy. Reviewers can override with one click and a recorded reason.
- Per-item refund paths in multi-item returns
- Recorded reasoning behind every recommendation
- Customer-facing language stays calm
Restock, outlet, repair, donate, liquidate — by case.
Routing reads condition grade, category, season, and partner capacity. Items move to the channel that recovers the most margin this week, not the channel that took them last quarter.
- Per-item routing inside one return
- Capacity-aware partner balancing
- Recovery rate audited per path
Product Issue Intelligence
Catch the cluster in week two, not season three.
Reasons cluster by SKU, color, and size. When a sizing or quality issue appears, merchandising sees it before the wave lands. The PDP fix that follows is small, fast, and measurable.
- Severity bands per SKU and reason
- Owner assignment and resolution tracking
- Slack and email alerts to the merchandising owner
Save the sale on size-related returns.
On size or fit reasons, the engine suggests the closest in-stock alternative — same product, one size up, in stock at the nearest node. One-tap exchange flow keeps the revenue.
- Reason-gated — never offered on quality complaints
- Honest stock awareness — never offer what we cannot ship
- Per-item suggestions in multi-item returns
The Monday email a returns lead actually reads.
Every Monday: total cases, recovery rate, fraud blocked, top reasons, top SKUs at risk, and the three things to look at this week. Built to be skimmed.
- Sent automatically per workspace
- Drill links into the underlying queues
- Slack mirror available
Wired into the systems your operation already runs on.
Verified at launch: Shopify Plus (Official Technology Partner), NetSuite (certified connector), ShipStation (webhook), and Stripe (direct API). OAuth credentials stored as workspace secrets, never exposed to the browser. Webhooks signed and rotatable. Other providers are on our roadmap and are not currently claimed publicly.
- OAuth and signed webhooks throughout
- Fail-Open mode if verified API availability drops below 99.9%
- Per-workspace integration health monitoring
Policy, automation, and access — in one matrix.
Role-based permissions, rule templates per brand profile, and an audit log on every change. Every applied rule writes an event your finance lead can read back later.
- Owner, admin, analyst, ops manager, CX manager, warehouse reviewer, viewer
- Saved rule templates per brand profile
- Audit trail on every rule change