Operations
10 April
Meet Max, an AI chat assistant for data analysis, powered by GPT-4
AnswerRocket launched the conversational tool to help CPGs zero in on insights held within growing volumes of data.

Photo by Pietro Jeng on Unsplash
AnswerRocket launched the conversational tool to help CPGs zero in on insights held within growing volumes of data.
For most businesses, any problems that arise with data are no longer tied to scarcity.
Enterprises have access to more data than ever, coming from a multitude of sources.
This can provide a path to enhancing understanding of consumers and how a business operates, all so that a business might improve.
But there’s now so much data that issues are stemming from abundance.
The question asked around CPGs and other enterprises is no longer, "Do we have data on this, and how can we get it?"
Instead, business leaders are asking, "What data do we have? And, how can we find it and put it to use in the most resource-efficient way?"
That’s an area where AnswerRocket is finding a role for the generative AI engine GPT-4, and a conversational approach.
The company recently released Max, an AI assistant that is designed to help businesses analyze their data through chat. On its recent launch, it was set to be used by CPGs such as Anheuser-Busch InBev and American Licorice Company.
The tool allows users to ask questions of the assistant that can be answered through their data. Then, they receive answers back in the form of insights and visualizations.
It arrives on the heels of the March release of GPT-4, which is the latest version of OpenAI’s large language model that powers tools such as the wildly popular ChatGPT.
AnswerRocket CEO Alon Goren has been working on technology that helps businesses find and analyze data since the company was founded in 2013, and before. Recent advances in large language models, particularly with the release of GPT-4, marked a step forward that enhanced both the technology’s ability to understand the question, and deliver answers to users that had the amount of detail necessary to solve a problem, Goren said.
These backend leaps have also improved the user experience. A tool like Max can now be approachable, even for those without data or tech skills.
“It's designed to be used as you would converse with an assistant,” Goren said.
Max aims to take high-level questions, and provide answers that go deep into a company’s data. Combining AnswerRocket’s augmented analytics platform with GPT-4, Max allows users to ask questions to the assistant in the same language they would use with another person when curious about a trend or outlier, and technical expertise isn’t required to dig into the data.
“It enables us to go from what the user said to a format where we can now query available content, search the data and then respond in a very natural narrative that a user can read,” Goren said.
While the questions can be basic and high-level, Max can provide statistical, diagnostic and predictive analytics that go deep into a company's data. In a large business, Goren said this can help advanced analytics spread across different departments. With an AI powered assistant, access to data analysis will be less tied to the decisionmaking that goes with expenditure and prioritization of human-powered resources.
“Suddenly, this long tail of products becomes much more manageable than it previously was with having to have a human in the loop to do every bit of analysis,” Goren said.
Along with availability, such a tool also holds the promise of making technology and data analysis more integrated into workflows than ever.
“A chat-based tool like Max can help more users feel comfortable interacting with data,” said Sabine Van den Bergh, director of brand strategy and insights for Europe at Anheuser-Busch InBev, in a statement. “Having an on-demand assistant that can quickly answer the questions that pop up throughout the day would enable our team to make data-driven decisions at scale.”
Nestle and General Mills joint venture Cereal Partners Worldwide, which is an existing AnswerRocket partner, sees the chat experience taking AnswerRocket "to the next level," said Chris Potter, global applied analytics at CPW.
"We need our teams to make informed, fact-based decisions," Potter said, in a statement. "Max will enable users across all levels of CPW to quickly access data and insights through intuitive questions and responses.”
AnswerRocket will be rolling out the technology to businesses throughout the second quarter.
Ask Instacart answers prompts with personalized recommendations.
A pair of recent launches from Instacart highlight how the grocery ecommerce company is integrating two of the key emerging areas of technology into its offerings: Generative AI and marketplaces.
Let’s take a look:
Instacart is seeking to harness generative AI to create a more personalized shopping experience.
A new tool called Ask Instacart that is launching this week is designed to allow customers to type in questions about specific recipes or general recommendations for an occasion. Embedded in the search bar, Ask Instacart also provides personalized questions to be asked by customers. In addition to specific items, it provides information about food preparation, product attributes and dietary considerations.
For those eying how generative AI will play a role in the shopping experience, Ask Instacart shows how search can be transformed into a place for discovery. Instacart is aiming to provide answers to the more open-ended questions that people would naturally ask, not just simply provide info in response to a question that has one answer. It shared the following sample prompts:
The tool is also showing the way for generative AI to integrate with retail media. Ask Instacart is designed to integrate with a brand's sponsored products campaign, so that the answers to questions that match consumer needs can also provide a way for brands to stand out.
To create the tool, Instacart combined the language understanding of ChatGPT with its own AI models. It added in catalog data from 80,000 retail partner locations around the country, which together have more than one billion shoppable items.
Beyond mission: Ecommerce marketplaces have honed a shopping experience where it’s easy to find what you’re looking for. But if shoppers want to happen upon something they didn’t know they needed, social media or the store is still the best place to visit. Instacart is showing how generative AI can make discovery a marketplace function. It also signals that advertising will come to generative AI by way of retail media. Going forward, the growth of discovery could make retail media more valuable as a tool for advertising that raises brand awareness, not just lower-funnel conversions.
Instacart will power a new virtual convenience store for the grocery chain Aldi.
Aldi Express will feature 2,000 of the most-shopped Aldi items, ranging from prepared food and snacks to grocery staples.
Drawing on 2,100 Aldi locations around the country, items will be delivered as fast as 30 minutes, the companies said.
“Through ALDI Express, we’re making shopping more convenient so you can satisfy a craving or get a missing ingredient in minutes,” said Scott Patton, VP of National Buying at ALDI, in a statement. “Together with Instacart, we’ll continue to find ways to innovate and make the online grocery experience even more effortless and accessible.”
Aldi began offering delivery via Instacart in 2017, and has since expanded services to include pickup as well as alcohol delivery.
Aldi’s marketplace moment? While Aldi previously offered delivery, making the assortment available through a virtual store offers the opportunity to create a marketplace for its goods. With the virtual store, it will more closely resemble DoorDash and Uber Eats, which have been expanding their grocery assortment. With a marketplace, additional revenue opportunities could open up for the grocer, such as advertising through retail media.