Lumtry · Agentic refund orchestrationRefunds decided by policy. Exceptions decided by you.
Lumtry decides refunds with deterministic policy plus AI reasoning, routes edge cases to your team for approval in the dashboard or Slack, executes the refund on Stripe or PayPal, and keeps a full audit trail.
Free plan with no time limit, or a 14-day Growth trial. No card needed.
Change a rule, watch the decision move. Runs in your browser, no sign-up.
Built by Agnotiq. Made for SMB e-commerce merchants: owners, ops leads, CX leads and finance.
Four ways refunds leak.
Start where it hurts most today. Every Lumtry module runs on the same guardrails, approvals and audit trail.
Refund decisions
A versioned policy with shadow mode and backtests. AI rationale on the gray areas, never above policy.
Explore refund decisionsReturns and labels
EasyPost and Shippo labels, QR and label-free returns, return-receipt gating and returnless refunds.
Explore returns and labelsDisputes and chargebacks
Evidence assembled from the case record, a narrative a person approves, and alerts before the window closes.
Explore disputesRetention and abuse
Retention offers before the refund goes through, Abuse Shield scoring and the opt-in fraud network.
Explore retention and abusePolicy decides. AI advises. You stay in charge.
Every refund takes the same path. AI reads the case first and hands the policy engine signals.
AI reads the case
A model reads each request first and hands the policy engine signals, including sentiment and high-risk outcomes.
Your policy decides
The published, versioned policy makes the call. Model output informs a decision; your rules make it.
A person takes the close calls
Held cases wait for approval in the dashboard, Slack, your helpdesk or by voice. Every step lands on the record.
Built so you stay in control.
Policy canvas
Your refund policy lives on a visual rule builder. Every rule is a block you can read, test and version.
Shadow mode
Run a new policy beside the live one and see where the outcomes differ before it goes live.
Approvals where you work
Approve in the dashboard, in Slack, in Gorgias or Zendesk, or by voice.
Nothing left half done
If a refund fails partway, Lumtry runs the compensation for the steps already taken and flags the case.
Held, never dropped
At your plan limit, new cases wait until you upgrade or the next period starts.
Automations
An MCP server for AI assistants, a public case API and signed webhooks.
“High-volume refund decisions made by policy you can defend, reasoning you can inspect, and approvals your team stays in control of.”
What it will never do.
Some lines are built into Lumtry, not left to settings.
Never override your policy
Model output informs a decision. Your published rules make it.
Never store card data
Lumtry keeps processor tokens only, never card numbers.
Never release a hold on model output alone
A held case waits for a person, whatever a model suggests.
Never rewrite history
Audit and webhook records are append-only.
What manual refunds cost you
A year of manual refund handling and abuse leakage, from your own numbers. They stay on this page until you open them on Lumtry.
About 160 refunds a month, 1,920 a year.
- Manual handling per year, 192 hours
- $4,800
- Abuse leakage per year, refunds that should have been denied
- $6,240
- Handling cost you could save, about 96 hours
- $2,400
Same method as lumtry.com: savings apply only your automation share to handling time. An estimate, not a promise.
Works with your stack, and your AI assistant.
Connects to
- Shopify
- WooCommerce
- BigCommerce
- Amazon
- eBay
- Stripe
- PayPal
- Slack
- Gorgias
- Zendesk
- EasyPost
- Shippo
- ShipBob
- ShipHero
Routing today across Anthropic, OpenAI and Mistral. Dashboard in English, French (Canada) and Spanish (US).
Let your AI assistant work in Lumtry
Lumtry ships an MCP server, an API and signed webhooks. Your assistant works under the same permissions and audit trail as your team.
MCP server docsStart free. Upgrade when it pays.
Lumtry is launching now. Start on the Free plan or a 14-day Growth trial, no card needed. Paid plan prices may still change, so the current ones live on lumtry.com/pricing.
Starter
Slack approvals and AI rationale for a growing store.
Growth
More volume and more channels for a busy support team. 14-day trial, no card needed.
Scale
Dispute automation and Abuse Shield included.
Enterprise
Custom terms, SSO and guided onboarding.
Lumtry questions, answered.
No. Your published policy decides. AI explains gray areas and can recommend, but a case outside the rules goes to a person.
Yes. The demo workspace has sample cases you can click through, and the Free plan lets you set up a policy before you connect anything.
Shopify, WooCommerce and BigCommerce stores, plus Amazon and eBay marketplace intake.
Stripe and PayPal. Store credit is issued through the store platform.
Yes. Shadow mode runs a new policy beside the live one and shows where the outcomes differ.
Lumtry runs the compensation for the steps already taken and flags the case, so nothing is left half done.
No. Lumtry stores processor tokens only.
The Free plan and the 14-day Growth trial are open now. Checkout for paid plans opens at launch, and current prices live on the Lumtry pricing page.
See Lumtry on sample data.
Change a rule, watch the decision move. Runs in your browser, no sign-up.
MarginTideWhat you chargeKnow every price. Protect every margin.
The sister productLumtry runs on the same patterns we build for clients: evals, guardrails and a person on every call that matters. A two-week Discovery scopes yours and ends with a fixed quote.
Book a fit call