Particular Audience, an AI-as-a-Service company, today announced Adaptive Transformer Search, an innovative search technology built using large language models and vector embedding. Adaptive Transformer Search promises to significantly improve online shoppers’ experience by dramatically reducing irrelevant or zero search results that plague e-commerce sites.
E-commerce websites lose an estimated $300 billion annually due to poor site search experiences, according to research from Google. Common problems include irrelevant search results that do not match customer intent and the abundance of “zero search results”. Particular Audience’s solution leverages breakthroughs in natural language processing and advanced AI algorithms.
“At Particular Audience, we’ve always focused on addressing the root causes of discovery abandonment with applied artificial intelligence”, said James Taylor, Particular Audience’s CEO. “With ATS, we’ve harnessed the power of Large Language Models, paired with our own vertical tuning to generate the most relevant search results right out of the box. No matter how niche or conversational a search is.”
Adaptive Transformer Search works by converting search queries and product information into dense vector embeddings that capture the meaning and context of words. This understanding of intent allows the technology to improve relevance and rank results based on their true similarity to the searcher’s need. Particular Audience has also pre-trained their models on retail data and products—known as Vertical Tuned Models—to generate optimized results from the start.
Particular Audience Provides a Fine-Tuned Version of Existing AI Models
The firm’s Vertical Tuned Models (VTMs) have demonstrated a significant boost in relevance compared to standard language models, Patrick DiLoreto, Particular Audience’s Head of Product, highlighted.
He emphasized that they have also incorporated a continual learning feature through Adaptive Reinforcement Learning, so the company’s models can be refined and improved over time based on click-through data and evolving customer preferences unique to each retailer.
Also read: Best AI Stocks to Invest in 2023
Adaptive Transformer Search’s benefits for e-commerce sites are profound. Compared to conventional keyword searches, the technology can:
- Increase search revenue by more than 20%.
- Reduce zero-search results by up to 70%.
- Enhance customer engagement through better result ranking.
Particular Audience helps e-commerce sites reinvent their search and discovery experiences through its AI-as-a-Service technology. Adaptive Transformer Search is the first solution to leverage the latest breakthroughs in natural language processing and vector search approaches specifically for the retail industry.
Most firms have been incorporating general-use LLMs and generative AI technologies such as ChatGPT from OpenAI and Claude from Anthropic. However, in the case of Particular Audience’s solution, the technology is designed specifically to aid e-commerce companies fine-tune their search capabilities.
The firm claims that its Vertical Tuned Models can generate a 37% improvement in relevance score rankings for searches compared to pre-trained open-source AI models.
This SaaS Has Raised Over $11M To Transform E-Commerce Searches
“We believe this revolutionary technology will not only transform the way consumers shop online but also set a new standard for search efficiency and customer experience in the ecommerce industry”, CEO Taylor further highlighted.
Particular Audience has raised more than $11 million from VC firms including $1.3 million in 2019 in its seed funding round and another $10 million two years later in its Series A financing.
According to data shared by the management team back in 2021, more than 50% of the firm’s revenue came from Europe. Particular had also surpassed $100 million in gross merchandise value (GMV) back then resulting in a 19x growth in just two years.
The Series A round was led by EVP and counted on other investors including Spykes, FiveV, and Carthona Capital. One of the firm’s most popular products is a browser plugin called SimilarInc.com, which searches the internet to find the best price for any particular product that a consumer is seeking to buy online.
Other Related Articles: