Kelley Toker:
Product Manager

How Transforms Shopping

Large language models (LLMs) have been around since 1967, with basic chatbots becoming common starting in 2011 and evolving into more sophisticated models by 2020. Currently, the most popular AI model, ChatGPT4, has over 1.7 trillion parameters enabling it to answer questions about everything from classic literature to writing SQL code. However, just like humans, ChatGPT4 is not an expert in all subject matters and cannot provide the level of detail that is commonly needed for consumer education. Enter 

Data shows that 96% of consumers research their big purchases online before buying. This means it is critical for brands to educate their customer base on their products so that their winning features and qualities can put them above the competition. But it can be hard to consume this knowledge via website. 

Specs often don’t have real-world meaning to the average person. For example, “8.4 inches of ground clearance and a 17.7° approach angle” means it is a great off-roading vehicle if it also has 4-wheel drive. But because consumers don’t often make this connection, it results in lost revenue and lower customer satisfaction scores. is working to bridge this communication gap and improve the customer web experience. is a conversation interface that can be deployed on a webpage to educate customers, simplify purchase decisions, and increase customer insights. It enables websites to provide a uniquely tailored experience to each visitor, ensuring potential customers are guided to the product features that matter most to them, which results in increased customer satisfaction and revenue. Under the hood, utilizes the conversational abilities of LLMs, such as ChatGPT, and pairs it with a customized network of datasets to accurately represent your company’s brand and product information.’s model is continuously improved by the conversations it has and updates with new product information as a company grows, “speaking” in a brand’s voice and bridging the communication gap. 

Question: “What lighting do you recommend for my kitchen?” 

Answer: “For kitchen lighting, it is critical to have a well-lit, low-shadow space that also allows you to accurately see the color of food. For this, I would recommend either halogen lights or LEDs with a high CRI value or a warm color temperature such as 3000K or lower. If you use an LED, you will need to have a diffusive cover over it to ensure minimal shadowing. Would you like to see a list of our available kitchen under-cabinet or overhead lights?” not only feels like a better experience to the consumer, but its success shows in the data as well. tracks bounce rate, funnel conversion, and user engagement so its impact can immediately be seen.

As continues to evolve, it will start incorporating individual user information, such as what they clicked on previously, to make recommendations. For example, if a customer looks at red car interiors before asking if a white crossover is available in their area, will check interior color too and let the consumer know if there are any white crossovers with red interiors available near them.

While there are many competitors in this space, is unique in how it builds a model for each website and manages the feedback cycle, making it a superior way to educate your customers.