Taste. Zero-shot. Discovery.

Solenya reorders your category pages and search results by each shopper's taste. It starts at their first click, and ranks a new product as soon as it's in your feed.

Live on MyRunway

One store, personalised for each shopper.

+27% revenue

Solenya: +27% revenue compared with MyRunway's own search and ranking. 90% interval +14% to +42%.

Measured in a randomised test on live traffic at MyRunway, a South African fashion retailer.
Read the case study →

How it works.

  1. Your product feed

    Solenya reads each product's photo and title and places it on a map of your catalogue.

  2. What each shopper views

    Every view pulls that shopper's taste toward products like it.

  3. Each shopper's page

    The page comes back in that shopper's order, inside your filters and rules.

Best for stores with more products than one page can show. A/B testing is built in, so you can measure it against what you run today.

Every product becomes a point.

In your feedPink sports braKept as filterspinkactivewearR 499In stock

Search finds what's nearest.

Try
Filter

5 matches.

  • Camel wool coat
  • Black quilted puffer
  • Olive hooded parka
  • Cream cable-knit jumper

A made-up store. The real map has hundreds of directions.

A shopper's taste is what they view.

The shopper's last 24 views0 of 24
Their taste on the map
Category page: starting order
  1. Black wrap dress1
  2. Camel wool coat2
  3. White leather trainers3
  4. Pink sports bra4
  5. Light-wash straight jeans5
  6. Silver sequin mini dress6
  7. Grey running trainers7
  8. Navy striped top8

Only product views count. The last 24 are kept across visits.

Search and similar productsThe words or the product on screen pick the matches. Taste can reorder them, if you switch it on.

Now be the shopper.

yourstore
Try
Click anything you like.The retailer's order, the same for every shopper.
SportswearEverydayCoatsEvening
  • Sportswear: Pink sports bra, Grey running trainers, Padded bib shorts, Pink seamless leggings, Blue ride jersey, White running jacket
  • Everyday: White leather trainers, Light-wash straight jeans, Navy striped top, Cream cable-knit jumper, Denim jacket, Black skinny jeans
  • Coats: Camel wool coat, Black quilted puffer, Olive hooded parka, Navy trench coat, Black tailored blazer, Yellow rain mac
  • Evening: Black wrap dress, Silver sequin mini dress, Pink satin slip dress, Silver strappy heels, Wine velvet midi dress, Black jumpsuit
Everyone starts with the same order, the retailer's, until their first view.Your tasteOn the first screen

What stays yours.

Solenya off for this shopper: your own order
  1. Camel wool coat
  2. Grey running trainers
  3. Black quilted puffer
  4. Olive hooded parka
  5. Navy trench coat
  6. Black tailored blazer
  7. Yellow rain mac
  8. Black jumpsuit

One request, start to finish.

Your storeCategory pages, search, product pages
  • Your pages and templates
  • Your filters and rules
  • The shopper ID, in a cookie or app storage
No login, no names.
SolenyaGoogle Cloud, europe-west4 (the Netherlands)
  • The product map
  • Each shopper's last 24 views
  • The tests you're running

Test it before you trust it.

A/B testing is built in. Split your shoppers between Solenya and what you run today, and the Console shows which earns more.

An example testSimulated. The Solenya Console draws yours like this.

Revenue per shopper, each groupSolenyaWhat you run today
Week 1Week 2Week 3Week 4
Solenya against today, with its rangeDifferenceRangeNo difference
Week 1Week 2Week 3Week 4
Early days swing both ways.

See it on your own catalogue.

Nine people, one jar.

Karl Hammerschmidt

Founder & CEO, MyRunway (South Africa's largest profitable fashion ecommerce site). 14 years scaling ecommerce.

LinkedIn →
Miguel Da Costa

First non-technical hire at Stitch (African payments infrastructure, $55M Series B led by QED). Scaled operations 0→150.

LinkedIn →
Nick Wilkinson

Nick publishes in Computer Speech & Language and at IEEE ICASSP, and leads the retrieval and indexing pipelines that sit at the core of Solenya.

Google Scholar →
Jean Durand

Jean graduated summa cum laude from Imperial College London, where he was awarded the CGW Schumann Prize and Medal. His research is in causal inference.

LinkedIn →
Preston Govender

Preston led the deployment of machine learning at scale inside one of Africa's largest banks. His work centres on the reliability and reporting of ML workflows in production.

GitHub →
Lawrence Godfrey

Lawrence comes from a software-engineering background in payments, where he worked on scaling highly secure workloads.

GitHub →
Anri Lombard

Creator of MzansiLM and MzansiText, contributor to popular open-source projects, and builder of services from first commit to production.

GitHub →
Bradley Mallett

Brad has led GTM strategies and helped to launch products in new regions, verticals, and customer segments. His focus is positioning Solenya as the future of eCommerce discovery.

LinkedIn →
Grethe Gultig

After a few years of in-house and agency design work, she spent four years designing dashboards for remote telecom and energy sites at scale.

Portfolio →

Hover over a pickle to meet them.

There's room in the jar. 2 open roles in Cape Town.