Book a call

EngineeringSoftware built to last

Custom software →Web applications →Mobile apps →SaaS product engineering → Cloud applications →AWS development →DevOps & CI/CD →Cloud migration →

IntelligenceAI in production, not slideware

AI & machine learning →Generative AI applications →AI automation & integrations →Data engineering & analytics →

Experience & GrowthDesign and growth, one roof

UX/UI design →Brand identity →Marketing websites →Digital marketing & growth →Creative & campaign design →
We ship our own productsSkypora, BroadKasto and Calynza, all live in production. All 17 services Ways to partner with us Insights
Case study · E-commerce · sample content

Recommendations that feel personal at scale.

Client
A D2C retailer (name under NDA)
Industry
E-commerce
Services
Generative AI · Data · Web
Timeline
10 weeks to launch
Team
AI engineer · 2 engineers · designer
The campaign dashboard
Before3,000+products, and a search that missed most of them
01 · The challenge

The catalogue outgrew its search.

Shoppers typed what they wanted in their own words and got nothing useful back. The team hand-picked "you may also like" lists that went stale within a week.

Approach0rebuilds: AI added inside the existing store
02 · What we built

An assistant that knows the shelf.

We indexed the catalogue, reviews and stock, and added an assistant that answers in plain language and recommends from what is actually in stock. Recommendations update themselves from what people browse and buy.

OpenAIRAGPythonNext.js
AfterLivein the store they already had
03 · What changed

Shoppers find it, and come back.

Search understands how people describe things, recommendations stay fresh on their own, and the team stopped maintaining lists by hand.

What we delivered

Ten weeks,itemised

Every engagement ends with a plain account of what was built and what changed. This one printed like this.

Samestore, no rebuild
AIassistant, live in production

SanctumCloud · Delivery

AI shopping assistant · sample receipt
Catalogue and data review1 wk
Search and recommendation models4 wks
Shopping assistant in the store3 wks
Testing with real shoppers1 wk
Launch and monitoring1 wk
Total to launch10 weeks
Thank you for building with us
“It answers like our best store assistant, at three in the morning.”
Head of E-commerceD2C retailer · sample quote
Next case
A lending dashboard that answers in seconds not spreadsheets.Read
Want a story like thisstart here

Your problem,our next case study

Book a strategy call Send a message