This is part of our #LaunchWithAI series. Interested in sharing your story? Contact us here!
One of the biggest challenges that AI researchers and developers face is in leveraging unstructured data. Unlike structured data, which is organized and easily searchable, unstructured data doesn’t have a pre-defined model or organizational system and is often difficult to analyze using traditional methods. However, as the amount of unstructured data in the world continues to grow exponentially, it’s essential to be able to extract insights and knowledge from this data.
Commerce.AI is equipping its customers to tackle the unstructured data challenge. The startup provides an AI-powered platform that helps businesses make data-driven decisions by leveraging the latest advances in natural language processing (NLP), computer vision, and machine learning technologies. The platform provides solutions that enable businesses to optimize their product data, conduct market research, and understand consumer feedback. With industry-specific taxonomy, customized dashboards, and customizable alerts, Commerce.AI provides insights and analytics to help businesses make data-driven decisions.
We sat down with Commerce.AI founder and CEO Andy Pandharikar for some insight into how his startup makes sense out of unstructured text, voice, and video data, and how working with Microsoft has given them more confidence in helping their customers.
Capitalizing on the value of unstructured data
Commerce.AI was founded with the mission to harness the power of AI to change the way commerce is done. Andy’s previous startup was focused on using data to help fashion retailers improve product sizing accuracy. After it got acquired, Andy and his team started exploring other scenarios in which unstructured data could produce valuable insights. And as he made clear, there was certainly no shortage of data to work with.
“There are 24 quintillion bytes of data created every day—that’s one billion billions,” Andy says. “And 90% of that data is unstructured.”
This unstructured data represents a potential goldmine of insights into customer sentiment, market trends, and usage patterns, as well as other information that can be used to make better decisions. However, unstructured data is harder to analyze than structured data, so companies often struggle to make good use of it.
“There is so much value out there for companies to derive if they can access and make sense of it,” Andy says. “When we ran that first startup, we saw that fitting is a huge problem in fashion. But then you start understanding it’s a data problem. How do you predict what type of clothing to produce so you can serve the maximum number of customers? Our data was predicting what size would fit.”
Gaining customers’ confidence with Azure and OpenAI Service
With his company’s mission to help businesses make data-driven decisions using advanced AI, NLP techniques, and machine learning algorithms, Andy is clear on what sets Commerce.AI apart from others in the field.
“We are an AI-first company,” he says. “How we differentiate from competitors is by centering our infrastructure around AI. We are focused on building and delivering the best experience to our customer—and that involves AI.”
“The large language models (LLMs) like OpenAI have gotten so good. I would advise other startups to not build their own models but find areas where they can adapt the existing models for the needs of customers.”
The company starts by defining a customer’s business problem, then develops a solution approach and tailors their AI infrastructure to suit those needs. Commerce.AI came to realize that as a startup, they wouldn’t be able to match the compute infrastructure of the big cloud companies. This led them to Microsoft Azure.
“When we started using Azure infrastructure, there was a tremendous amount of reliability when it came to safety, scalability, and compliance,” Andy says. “We noticed it gave our customers confidence.”
And Azure provided more than just infrastructure. As part of Microsoft for Startups Founders Hub, Commerce.AI joined the Azure OpenAI Service (AOAIS) beta program after over a year of testing Chat-GPT3 internally. The service allowed them to provide generative insights to their clients in place of tables or text, and they didn’t have to program it from scratch. For example: perhaps you want to summarize a lengthy customer service call with the goal of producing better scripts for agents to reference in the future. Azure OpenAI Service could look at that summary and other past scripts and develop something that resonates with customers.
“We’re an applied AI company, so when it comes to solving a problem, we always look at what’s out there and open. There are outstanding AI tools and infrastructure out there to solve the problems our customers need to address. The partnership with Microsoft gives us tremendous confidence to take those features to our customers. So when we talk about trust, safety, compliance, and scalability, it’s all about how they’re powered by Azure OpenAI Service.”
Andy’s advice to other startups? Don’t try to rewrite the book when it comes to AI.
“The large language models (LLMs) like OpenAI have gotten so good,” he says. “I would advise other startups to not build their own models but find areas where they can adapt the existing models for the needs of customers.”
There have been significant advancements in the field of LLMs in recent years, especially since the introduction of transformer-based architectures such as BERT, GPT, and T5. These models have pushed the limits of natural language understanding and generation, leading to several breakthroughs in language-related tasks, including pre-training, transfer learning, multimodal learning, and interactive models.
There are hurdles to clear for startups looking to utilize LLMs, such as the vast amount of high-quality data needed for them to function effectively, but that’s why Andy suggests leveraging pre-trained models.
“Spend your time and resources using OpenAI infrastructure in an effective way for your customers rather than trying to rebuild,” Andy says, “The OpenAI models are really great. You can’t beat them.”
It was clear from our conversation that Andy saw tremendous value from not only Azure OpenAI Service but also membership in the Microsoft for Startups program. Andy expressed his enthusiasm for the infrastructure and strategy that the Azure team and Founders Hub provides.
“At the early stage, when you’re deploying and trying different strategies and doing free POCs, cloud credits help,” Andy says. “But when you start getting the product-market fit, you also need to have a good go-to-market strategy.”
Working with Microsoft go-to-market experts, sales, and customer success team members helped Commerce.AI home in on their strategy, Andy says, giving them momentum and a path to scale. And since many of their customers already have their own cloud infrastructures on Azure, it’s easier for Commerce.AI to deploy their solutions.
For more tips on leveraging AI for your startup and for access to Azure’s AI services, sign up today for Microsoft for Startups Founders Hub.
