Leagăn - Reviews and UGC Platform with Built-In Safety Escalation
Honest Reviews for a Baby Products Retailer
Designed and built a clickable reviews and user-generated content platform for a Chișinău baby and nursery retailer selling in Moldova and Romania: a moderation queue that works to a written policy, review detail with personal data masking, and a safety escalation desk that picks up any review describing a hazard to a child.
Most of our work is under NDA. Per agreement, we can share the work but not the company details - all names and identifying information shown here have been changed.
Impact
Dashboard, moderation queue with its quick view panel, review detail, safety escalations and safety case detail. Approve, mask, reject with a reason, escalate and send to safety all update the record as you click.
A review, question or photo that describes an injury or near miss opens a safety case in parallel with moderation. Cases are typed by 10 hazard types, and any case past the target turns red in the sidebar and on the dashboard.
The dashboard compares the rejection rate of 1 and 2 star reviews with 4 and 5 star reviews and warns above 1.5 times. The 8 rejection reasons contain none for a low rating.
Reviews in Romanian and Russian with filter hits, fake review signals and decision history, plus 12 safety cases with timelines, 14 products, 8 brands, 18 shoppers and 8 staff users.
The Problem
Codrița Baby's reviews ran on a plugin bolted onto its old shop. Anyone could post without buying, so a pram launch drew a burst of five-star reviews from new accounts while a competitor's page collected one-star reviews from people who never owned it. Reviews were approved by hand from an email inbox with no written rules, manufacturer reviews were copied in without a label, and twice a parent described a real hazard that sat in the inbox for a week before the product team saw it.
Our Solution
Leagăn takes a review from submission to publication under a written, versioned policy. Filters flag profanity, personal data, links, safety keywords and fake review signals, moderators decide each review with a named reason, and anything that describes harm to a child is escalated to the product safety team at once, with a timeline of who was told what and when.
The Challenge: Reviews Shoppers and Regulators Can Trust
Codrița sells prams, car seats, cot mattresses and toys in Moldova and Romania, where a faulty product can hurt a baby. Its reviews have to be honest under EU consumer review rules, safe for children shown in photos, and fast enough to catch a hazard before another family meets it.
Honesty Rules on Every Review List
Since the Omnibus Directive amended the Unfair Commercial Practices Directive, a trader showing reviews must say whether and how it checks that they come from real buyers, and must not publish only the positive ones. Codrița applies the same standard on its Moldovan storefront as in Romania.
Hazards Hidden in Ordinary Reviews
A toy whose eyes come loose or a cot mattress that does not fit snugly is reported first in a review, not to customer care. Under the GPSR a distributor that learns of an accident must inform the manufacturer and, where the product is dangerous, the authorities.
Fake Reviews Without Punishing Honest Ones
Brand-new accounts, shared devices, bursts on one product and copied text point to bought or coordinated reviews. Each signal on its own is weak, so automation can hold a review for a person to check but cannot remove it.
Children and Personal Data in Content
Parents write phone numbers, addresses and national ID numbers into reviews and post photos of their babies. Leagăn has to mask personal data under the GDPR and Moldova's Law No. 195/2024, strip photo location data and treat details of a child's injury as restricted.
Our Approach: From Review Rules to a Clickable Platform
Spec and Rules Research
Mapped 10 roles, from moderator and product safety officer to brand user and shopper, 38 screens across staff, brand portal and mobile-first shopper views, and the data model for reviews, media, questions, decisions, statements of reasons, safety cases, signal clusters, investigations and syndication sources. Wrote down the rules the design follows: the Omnibus review rules, the GPSR, the Digital Services Act, the GDPR and Moldova's personal data law.
Clickable Demo on a Production UI Kit
Built 5 key staff screens on a production dashboard kit: the dashboard with the safety strip, KPI cards, rating mix and queue health by age band, the moderation queue with highlighted filter hits and keyboard shortcuts, review detail with order proof and signal breakdown, safety escalations with products that have 2 or more cases in 90 days pinned, and the safety case with its timeline and decision record. All actions run on local state.
What a Production Build Adds
Integration with the order system for verified purchase, Brevo or SendGrid and Viber via Infobip for review requests, Cloudinary or Mux for media, image and text classifiers, Microsoft Presidio for personal data, DeepL translation, Bazaarvoice or brand feeds for syndication and EU Safety Gate for recalls. Staff auth by role, a tamper-evident audit log, and the remaining screens: media and Q&A queues, integrity, review requests, reports, the brand portal and the shopper storefront.
Why It Works: Moderation That Can Explain Itself
Every Rejection Has a Named Reason
Rejecting a review means choosing a reason from the versioned policy, with its shopper-facing wording in Romanian, and a statement of reasons is sent. Bulk approve only covers rows with no filter hits or signals, and there is no bulk reject.
Safety Runs Beside Moderation
A review that answers yes to the safety question or hits a safety keyword opens a case at once, so the safety officer starts while the review is still in the queue. A safety review is never held back for being negative; it is published once personal and medical details are masked.
Personal Data Masked, Not Removed
Filter hits are highlighted in the text with the rule and confidence, and Mask and approve publishes the review with the phone number hidden while the original is kept for audit. Parent contact and medical details on a safety case are restricted, and opening restricted media is audited.
Low Ratings Stay Visible
The dashboard sets published reviews by star next to rejected reviews by reason, with the rejection rate for each star rating, so any pattern of removing bad reviews shows up on the first screen anyone opens.
Related case studies
View all: E-commerce & D2C
Nakhuda - Bilingual AI Shopping Assistant Grounded in the Product Catalogue
Designed and built a clickable console for an Arabic and English AI shopping assistant at a Kuwait City electronics retailer: every product claim cited to a catalogue field, compatibility decided by fixed rules, nothing added to a cart without the shopper's tap, and a review queue for every suspected invented fact.
Tartera - Headless Storefront and Merchandising Admin for Four European Markets
Designed a mobile-first headless storefront for an Andorran outdoor apparel brand selling to Andorra, the EU, the UK and Switzerland, and built a clickable demo of its staff admin: storefront health, merchandising rules, a collection grid editor, CMS content status and per-market prices, taxes and legal settings.
Enterprise Software Engineering - Human + AI
Contact Us
© 2026 - Copyright. All Rights reserved Oceanlab
Location
Global Headquarters
21 Merchant Road, 3rd Floor, Singapore 058267
Main corporate office and leadership operations
North America Office
548 Market Street, PMB 12345, San Francisco, CA 94104
Client operations and business development
India Office
402 Shivalik Shilp II, SG Highway, Ahmedabad 380015
Engineering and delivery center
Enterprise Software Engineering - Human + AI
Contact Us
Locations
Global Headquarters
21 Merchant Road, 3rd Floor, Singapore 058267
Main corporate office and leadership operations
North America Office
548 Market Street, PMB 12345, San Francisco, CA 94104
Client operations and business development
India Office
402 Shivalik Shilp II, SG Highway, Ahmedabad 380015
Engineering and delivery center
© 2026 - Copyright. All Rights reserved Oceanlab