Transforming Uber’s Mobile Experience with APIs, AI, and ML

By now we all know that mobile communications and “chat-style” apps have effectively become the browser within which billions of people (many of whom have never before owned a computer) will access the web. Asian chat apps like WeChat and LINE, among have pioneered innovation in this sector, becoming much more than just an “app,” but transforming into a platform to handle payments, connect with brands, transact with businesses, and much more. The moves we see inside Facebook Messenger angling toward payments, a conversational assistant (Moneypenny), and talking to local businesses are just an inkling of what we can expect in the future. It’s hard to distribute mobile apps, and there’s a power law among the apps we actually use (mostly communications-based), we have the text message entry bar as the new command line interface.

Most people think of “chat apps” when they think of this, and that makes sense. We are using iMessage with our friends, texting on Messenger, and talking to friends overseas with WeChat, etc. It’s easy to imagine each of these services offering conversational assistants, powered either by humans or machines, to help us transact or get things done. Why should we leave Messenger when we’re coordinating with a friend about getting dinner, when the system can just know what we want and respond to our requests implicitly and book a table for us via the OpenTable API?

Now, I think another popular mobile app will also transform into having these characteristics — Uber.

When it started, the Uber app was about calling a black car. Then, it evolved into offering us UberX, and most recently, UberEATS. And, they’re not done, with reports dripping out they’re experimenting with commuters, commerce, group rides, and more, I believe we will see the Uber app transform in ways that will be hard to imagine today. For example, there may be a search box or text box to enter information and/or chat with drivers and friends, and in those conversations, the text box becomes another command line interface ripe for bots to interact with. An example: You call an Uber to meet a friend at a restaurant, you can deep-link out to OpenTable to reserve a table, or maybe just text within Uber you need a table at Restaurant X and it takes care of it in the background. even to the point where your credit card is used for checkout (like inside OpenTable right now).

There are many examples herein. The larger point is that just like our chat apps are transforming into colossal platforms hooked into myriad services, Uber’s mobile app won’t just be a place to order your car or lunch…they will want to capture more of your attention and more of your intent, and now that more AI-based, NLP-based, and ML-based technologies are accessible to third parties, we should expect Uber — like Messenger — to want to hook us into a platform where we can search for, discover, communicate with, and transact with a host of local stores and service providers. These apps will pick up on the natural language we use, they’ll leverage artificial intelligent bots which can parse implicit signals and machine learning algorithms which can learn our preferences over time and either compress transaction time or help us discover new things. For many people who live in cities, Uber is a homescreen app, and it will be fascinating to see how it evolves from EATS to commerce to other areas.