Signal AI uses AWS to Help Brands Understand What the World Thinks

Signal AI was founded in 2013 when its founders realized that brands needed a way to manage and interpret vast amounts of data. “Just as the explosion in digital data was taking off, we realized there was more data out there than any human can consume,” says Luca Grulla, chief technology officer (CTO) at Signal AI. “And that’s where artificial intelligence and machine learning help.”

Today, the company has more than 700 customers, including Bank of America, Deloitte, and Google. It was recently named one of the UK’s fastest-growing technology companies in the Sunday Times Sage Tech Track 100 league table, and was a 2022 winner in the Rocketship category of the AWS Software Startups Awards. “We’re an external intelligence platform supporting decision making,” says Grulla. “We make sense of what’s being written about a brand and help it make better decisions by presenting all the data in an understandable, digestible format.” Using AWS, Signal AI’s platform runs 24/7 and processes 3–5 million documents daily.

Signal AI has been using AWS to run its infrastructure and all its services from the time it was founded. In the years since, the company has added new AWS services. Signal AI has the on-demand database service it needs through a mix of Amazon Aurora, for unparalleled high performance and availability at a global scale, and Amazon DynamoDB, for single-digit-millisecond performance.

Signal AI’s data lakes are built on Amazon Simple Storage Service (Amazon S3), which is built to retrieve any amount of data from anywhere in the world. Today, the company uses dozens of different AWS services, ranging from cloud infrastructure to translation.

Signal AI works closely with AWS engineers to identify new services that could improve its offering. “AWS might be big, but it’s a great company to work with for a scaleup like ours,” says Grulla. “We have a great relationship with AWS—we’re able to bounce ideas off them. They know what they’re talking about and will always tell you the right thing to do, and not sell you something you don’t need.”

Signal AI’s platform needs to work fast given the volume of data it processes—a throughput of up to 5 million documents daily. In just 25–30 seconds, it can transform a text document into valuable, contextualized insight. The AI is smart enough to know the difference between Apple (the company), from apple (the fruit), or Fiona Apple (the singer).

Using AWS, Signal AI’s platform can achieve the highly sought-after P90 document processing time of below 30 seconds. The cloud processing required is high, but AWS provides the speed and processing power that Signal AI needs to hit the crucial P90 target.

The move to Amazon Translate means that Signal AI can process and translate 75 billion characters per month. “Signal is in the business of ML and AI, not in the business of building data centers,” says Grulla. “With AWS, we can scale horizontally with ease. If we have a sudden spike in customers or processing demand, it’s no problem. We also have a faster time to market through easier prototyping. And by using reserved instances on AWS, we’re saving 30 percent every year compared to on-demand pricing.”

Beyond scalability, experimentation and innovation are important benefits for Signal AI. “Innovation needs to be constant,” says Grulla. “You don’t have the luxury of disappearing from the market for a year or two to work on your product. Innovation needs to be rapid, iterative, and conducted in the flow of work.”

Using AWS, Signal AI can scale quickly and experiment with different models and datasets, meaning faster prototyping and shorter time to market for new features. It means better services and experiences for Signal AI’s customers.

Grulla is clear on the advantages that Signal AI has gained by using AWS. “AWS has given us the scalability and flexibility we need,” he says. “They’ve been with us from the start, helping us grow from startup to scaleup, all while letting us focus on our core offering. And using AWS gives our customers confidence—using a globally recognized cloud infrastructure service tells our customers that we take things seriously.”