Background

Back in 2018, Google started using neural networks to understand search queries better. Then, in 2019, they made a big improvement with something called BERT, which helped them understand the meaning behind different word combinations.

Fast forward to now, and the big deal is these super-smart language models (like Google's Gemini). They've advanced so much that we can have conversations with chatbots or type queries into search bars like we're talking to a personal assistant. You might have seen this in action on search engines like Bing's Copilot or on some brand websites and apps.

Google - Gemini

Google's latest model, Gemini 1.5 Pro, is even more impressive. It can understand a lot more context (like up to 700,000 words), and it's not just limited to text—it can handle video and audio too. Plus, it can follow along in a conversation, like when you're chatting with a virtual shopper about your preferences and history. This means it can give more detailed and relevant responses without needing as much fine-tuning or specific facts.

But there are some concerns about using these models, especially when it comes to ethics. Some worry that relying on them too much could lead to mistakes or biases in search results. Google recently introduced some AI-generated search features in the UK but there's still debate about whether this is the right move, considering the potential for errors and other issues.

Present day - Google Vertex

Google's Vertex AI Search, launched in summer 2023, enables businesses to create smart search features using the Gemini model. It's like having a supercharged search engine that understands your company's needs. Think of it as combining a really smart search tool with your own company's knowledge.

Earlier this year, Google Cloud introduced a special tool for retail called conversational commerce, powered by this technology. Now, businesses can set up virtual shopping assistants in just a few weeks. These assistants can help customers find products they'll love based on their interests. Retailers can also improve their online stores with top-notch search, browsing, and product suggestions that are seamlessly integrated. They can even fine-tune the system to match their unique products and how customers search for them, making it easier to show the right products to the right people.

But Google isn't the only player in town. Walmart, for instance, has teamed up with Microsoft Azure OpenAI Service and developed its own tools. And there are other companies, like Coveo and Algolia, offering similar solutions for commerce search.

If capabilities like this are of interest to your organisation, please contact us now for a no-obligation discussion.