veridive is now an applied AI company. Looking for the answer engine?Looking for the answer engine? What happened

veridive TR Start a project Menu

Work Illustrative engagement · Commerce intelligence

Returns that resolve themselves. Almost.

Client (illustrative)
A multi-brand home & living retailer · 120 stores and online · Türkiye
Scope
Discovery Sprint Pilot to Production · 10 weeks · Team of four
Fig. 5.0 — Four channels, four rulebooks, one reviewed decision

01 Key results

Key results

Illustrative pilot targets

Return decision time

3 days to same day

Cases needing full manual review

100% to about 35%

Policy consistency across channels

Measured weekly

02 The situation

The situation

Returns arrived through stores, the website, marketplaces and the contact center — each with its own form and its own rules. A small team reviewed every case by hand: reading the customer’s message, checking the order, looking at photos, and applying a policy that had grown to forty pages. Decisions took days, and the same case could get different answers in different channels.

03 The question

Could a system prepare each return decision — with the evidence attached — so the team only had to confirm, adjust or escalate?

04 What we built

Evidence first, then a person decides.

  1. A

    A triage assistant that reads the request, pulls order and product data, and checks photos against the product record.

  2. B

    A policy engine that cites the exact paragraph behind every recommendation.

  3. C

    A review screen where the team confirms, edits or escalates in one click.

  4. D

    Weekly quality reports comparing recommendations with final decisions.

  1. 01Order & product data read-only
  2. 02Photo check
  3. 03Policy retrieval with citations
  4. 04Recommendation + confidence
  5. 05Human review
  6. 06Decision logged
Fig. 5.1 — The returns flow; a person decides at step 05

05 How we worked

Ten weeks, as a stream.

The band widens where the work was most intense and settles as ownership moves to the team.

  1. Weeks 1–2

    Align & validate

    Walked the returns process in two stores and the contact center. Built an evaluation set from 400 past cases. Agreed acceptance criteria.

  2. Weeks 3–6

    Build

    Weekly demos with the returns team. Integrated order data and the policy library. Designed the review screen with the people who’d use it.

  3. Weeks 7–8

    Pilot

    Ran alongside the existing process on live cases and compared decisions daily.

  4. Weeks 9–10

    Embed

    Trained the team, wrote the runbook, set up monitoring and a monthly review.

06 What changed

Evidence attached. Judgment where it counts.

  • Most standard cases now arrive with a recommended decision, the evidence and the policy paragraph behind it.
  • The team spends its time on unusual cases: damaged goods, repeat returns and high-value items.
  • Customers get an answer the same day in most channels.
  • Store managers see the same policy applied everywhere.

07 Guardrails in this project

Guardrails in this project

  • Refunds above a threshold always need a person.
  • The system never contacts customers directly.
  • Photos are processed in-region and deleted after the decision.
  • Every recommendation shows its sources.

08 What’s next

Extending the same approach to warranty claims and marketplace disputes.