How AI Can Bring a Personalized Experience to IoT
Let's look at the synergy that exists between AI and IoT solutions when it comes to personalizing your users' CX, using Neura as an example.
Join the DZone community and get the full member experience.Join For Free
Sometimes it's easy to become a tad jaded about all things IoT, particularly in the consumer market due to the slew of single-purpose capabilities, copycat products, poor interconnectivity, and of course, the litany of security embarrassments. Then you meet a company that offers a way to make connected products personal, instinctual and truly targeted to the individual. I've been following the work of Neura, for a while and I met with their CMO, Kris Bondi, at Mobile World Congress in Barcelona to get an update on the company's progress and gain further insights.
Effectively a B2B2C company, Neura has created an AI engine that turns IoT environments into connected universes that allow companies to connect with their customers during the most meaningful moments. Neura enables consumers to take control of their smart home devices – from Amazon Echo, Nest thermostat, Hue Lights, Ring smart doorbell, refrigerators and more – and make their smart homes more intelligent with the integration of true AI. Their AI engine integrates with multiple data channels to provide situational awareness for individual customers, ensuring personalized, highly relevant engagement. Their technology utilizes smarter integrations through machine learning to enable users to get to an unprecedented level of technology personalization. Apps and devices are enriched with insights not only about users’ past and present actions but also calculated predictions about the next thing they’re going to do.
How it Works
Neura's tech relies on the reality that the users of IoT tech are surrounded by connected devices throughout most of their day, whether WiFi at work, a smartwatch, a Bluetooth headset, or their connected car. Because of this, an endless flow of connections with their specific patterns and clusters is exposed through predefined moments. Neura's cognitive computing techniques find the patterns that define the users’ daily lives to detect and understand the relations between users and their surroundings, their work and home routine, exercise habits, etc. and provide companies with a simple to consume API.
I asked Kris about how it works and she explained:
"The Neura SDK is integrated into an app and that starts pulling the data from the phone’s sensors as well as wifi and Bluetooth signals. This then feeds into the AI engine. It's a hybrid AI engine, therefore the majority is cloud-based and a lightweight version is in the SDK. From this, we create a digital double. Our customers’ products access specific insights and predictions via API calls like the user has gone running or user has left home. The lightweight AI engine is able to react in real time to an anomaly and change the consequential actions. For example, if you usually go to bed at 10 pm and tonight you’re awake, drinking in a bar, the things that usually happen (thermostat change, door locking), aren’t going to occur because you’re not where you’re usually are and you don’t necessarily sleep with 85 people around you."
Neura has grown their customer base as well as providing greater service benefits for the end users. Kris said:
"We are increasing significantly each quarter. Monthly active users are those that have done API calls. Our revenue model is based on monthly licensing fees for active users. We have significant penetration in the apps, IoT devices around smart homes are integrating and starting to roll out. Our biggest markets are digital health, insurance, and smart home as well as transportation and travel which are starting to grow swiftly."
More recently, Neura moved into connected cars, enabling parking, airline and commute apps to interact with users in the moment. In terms of B2B, this means that companies launching upgraded connected cars and transportation apps can fully understand who each user is, then anticipate and adapt to their needs throughout the day. This data is critical for apps and vehicles to respond to physical world variables such as a driver that has had little sleep or a carpool with several people with very different music tastes.
Neura in Action
I was interested to learn how an app embedded with Neura's AI would differ to one without. Kris gave the example of a fem tech app called My Days tracks and predicts periods, ovulation and fertility. For women trying to get pregnant, temperature monitoring is part of the process. Kris explained that with temperature monitoring the accuracy diminishes after 10 mins of waking up. If you have a setting for 7 am and wake at different times you may have missed your window, so accuracy is important. Before rolling out to their full user bases of several hundred thousand, My Days tasted Neura’s AI on 80,000 by rolling out Neura's AI on half of these devices’ alert notifications. The company found that alert acknowledgment and engagement in the app's user with Neura's AI increased 10X and they've since rolled it out to all their Android users and are planning to do so with iOS soon.
Neura will continue their journey to make IoT and mobile apps more meaningful for the end users. They're focused on improving their customers' APIs and will soon be announcing an engagement API where customers can see context around user engagements. We were always promised that IoT would make our lives just not connected but easier and more meaningful and it looks like thanks to Neura, that promise is becoming a reality.
Opinions expressed by DZone contributors are their own.