You may not have heard of Ada yet, the Canadian company revolutionizing the customer service experience with AI, but with household brands including Meta, Verizon, and AT&T among its 300 customers, you’ve probably interacted with its technology without realizing it.
Since being founded in 2016, Ada has become a leader in AI-powered automation, and when it comes to generative AI, it’s well ahead of the game. Through an omnichannel automation experience, Ada's mission is to solve the most – and the most complex – customer inquiries with minimal effort.
There is a tremendous amount of unstructured data, with a range of quirks, peculiarities, and data formats across many channels that a large language model, or LLM, needs to incorporate in order to generate a constructive response or solution.
In order to move quickly and securely, this is an extremely important consideration for companies building AI solutions using customer data. Mike Gozzo, Ada’s Chief Product and Technology Officer, joined MongoDB’s Chief Executive Officer, Dev Ittycheria, and Chief Product Officer, Sahir Azam, on stage at MongoDB’s 2023 .local NYC keynote to discuss how Ada is able to innovate faster and more confidently, staying ahead of the generative AI curve on MongoDB Atlas.
From the beginning, Ada wanted to get solutions to market as quickly as possible. In order to do that, it needed a multi-cloud no-SQL database platform that enabled innovation without constraints. As soon as MongoDB Atlas became available, Ada quickly moved onto it, because as Gozzo shared, “We wanted to focus on our business of automating customer service inquiries, not on administering Mongo. This was a huge ‘unlock’ for us in terms of velocity.” MongoDB Atlas underpins all of Ada’s solutions, harmonizing multiple data types and enabling the team to focus on exploring generative AI and large language models.
“Using MongoDB means we’re not limited in how we source data if we want to build something new. We can query unstructured data and use it to train other models,” explains Gozzo. “We use generative AI effortlessly throughout our product stack to automate queries and provide support that goes beyond just answering multi-step queries. With MongoDB, we’re able to ship new products in just a few months.”
Gozzo goes on to share why Ada has continued with MongoDB since the startup’s inception: “Having the flexibility and ability to just pivot on a dime was really important. We saw that as we advanced the company and brought in new channels and new modalities, having one data store that can be easily extended without crazy migrations and that would really support our needs was absolutely clear from MongoDB. We’ve always stayed the path with Atlas because the performance is there, the support from the team is great, and we believe in having less dependency on one central cloud vendor that MongoDB allows us to have.”
Ada recently adopted Change Streams to iterate on, and train models in real-time while keeping the internal data lake in sync. This empowered the team to build a distributed event processing system that powers bots and analytics.
Ada’s also exploring Queryable Encryption, which helps to advance AI training while keeping conversations private. “What’s more interesting than just throwing transcripts into an AI model is to distill the core message and guidance behind encrypted conversations. It can look at the feedback from the bot manager on quality and learn what good interactions look like,” explains Gozzo.
In a few years, AI will likely be embedded into every app we use. With MongoDB, Ada is one of the market disruptors taking us there. “We’re super excited to be working with MongoDB Atlas, and can’t wait to see what else is in store,” concludes Gozzo.