Thursday October 1, 2020…London,UK Ecommerce search, personalisation and merchandising technology provider, Attraqt, (LON: ATQT) has today announced it has entered into an agreement to acquire Artificial Intelligence (AI) Search IP assets from Aleph One GmbH.
This acquisition transforms Attraqt’s ability to deliver proprietary AI-driven search capabilities to over 300 global brands across its customer base. The acquisition follows a five-month partnership with Aleph, which on-boarded a select group of customers including global beauty and fashion brand, PrettyLittleThing and a major UK supermarket to incorporate the Aleph self-learning product search algorithms for complex search queries – achieving immediate impact on KPIs and incremental revenue*
“Brands need to step-up and respond to commercial pressures and complex shopper expectations. Our investment in search engine artificial intelligence is a direct response to heightened expectations from consumers who are used to sophisticated search experiences on Google or Amazon. This acquisition immediately empowers our customers with a new, AI-powered ecommerce search engine and a proven ability to deliver contextually accurate product relevancy. This allows us to serve our customers with advanced, ‘Google-like’ search experiences across their digital channels, whilst unlocking previously dormant revenue opportunities, said Mark Adams, CEO Attraqt.
I’m thrilled it also means that following our 2019 acquisition of the Early Birds personalisation platform, Attraqt remains steadfastly committed to fast-tracking our transformation into an AI business, alongside the Aleph-One R&D team.”
Traditional search engines rely on concepts that are not capable of making a complex match between consumer context and product information. This requires a rich understanding of the shopper, as well as of the product - including image, attributes, or text-based information. While AI is often used to solve the issue by improving the ranking of a product, its value is compromised if the core search engine delivers only limited or zero results*.
The algorithms are based on a foundational technology developed at Aleph-One: the‘ product fingerprint’; this technology scans and comprehends all available data about a product: all images, textual descriptions, categories, labels, textual attributes, prices and other meta data. Based on combining natural language processing, deep learning and computer vision the algorithms build a comprehensive semantic representation of products and search queries. This allows the AI to understand and continuously learn underlying customer intent based on search requests, and match relevant products, even for products whose attributes are not explicitly mentioned by the customer in the search query. The search engine also caters to the growing demand for natural language or voice search requests.
Aleph-One developed their “product fingerprint” algorithm to counter commercial losses brands face due to “low confidence” searches which deliver poor recommendations and lost opportunities for a brand to engage and convert shoppers. The Aleph-One R&D team will work alongside Attraqt and further develop an AI IP portfolio on top of Attraqt’s orchestration platform. The Berlin-based company will also continue to operate and trade under its registered name.
“This is a game changing moment for Aleph-One and Attraqt,” says Aleph-One’s managing director Duncan Blythe. “With this IP acquisition we have the opportunity with Attraqt to bring our breakthrough technology and algorithms to the biggest online brands in the world, allowing for more nuanced and enjoyable user experiences which interface directly with brand’s websites.
It’s exciting for us to join forces with an Attraqt R&D team committed to pushing the boundaries of how AI can impact entire customer journeys, whilst at the same time continuing to forge new innovations in AI which will flow directly into the Attraqt product offering.”
Attraqt will continue to roll out its suite of AI products to further transform search as well as recommendations and personalisation capabilities over the next year.
* For example: A shopper creates the following search query: “leopard summer dress” but the word “leopard” is not part of the text description of the product. The Attraqt core search engine will find and display products with leopard patterns even if these are not in the product descriptions based on a semantic representation of the product. The search engine will also recommend similar products from the catalogue, where traditional search engines may present a very limited result set or zero results.
Attraqt powers exceptional shopping experiences for over 300 of the world’s leading brands, manufacturers and retailers. Attraqt provides a set of API-enabled, algorithm-driven, intelligent SaaS services covering personalisation, search, navigation, merchandising, recommendations and internationalisation.
The platform unifies and empowers all key customer journey stakeholders. Enabling business, data and technical teams to successfully leverage existing ecommerce technology investments. Ultimately ensuring individualised and connected micro-experiences that support commerce discovery through to inspiration and purchase, and beyond.
The Attraqt platform orchestrates AI models in real-time while allowing clients to integrate their own algorithms. This helps build true personalisation strategies and create differentiated experiences at scale, so our customers can exceed the expectations of today’s shopper, and still reach commercial goals.
Attraqt is publicly listed on LSE: AIM and headquartered in London, with offices in Amsterdam, Chicago, Paris, Sofia, Hamburg and Sydney.
"Aleph-One GmbH is an AI innovation lab based in Berlin Germany, which develops proprietary and open-source AI software and technology in line with emerging trends and discoveries from the research community. In 2019 Aleph-One brought its first product to market: Aleph Search. This is a complete search suite for e-commerce, comprising semantic text-search, visual recommendations, autocomplete functionality, reverse-image search and semantic business rule design."