The only living creature in Steven Hansen’s life that does not use the Amazon Echo is his dog.
“I use it both at work and home. My wife listens to music on it, and my son asks Alexa for help on his math homework,” said the chief technology officer at Rain, who leads the agency’s 20-person voice-experience team from its Saratoga Springs, Utah, office.
Voice has become a huge focus area for Rain ever since the agency debuted its recipe library for Campbell’s on Amazon Echo last year. It has not only developed Echo skills for a range of clients since but also recently built a smartphone app called Reverb, which lets anyone interact with Alexa — even if they don’t have an Echo device.
Hansen, who started off as a developer before joining Rain in early 2011, said he has come to realize how much more important content and tone is when it comes to designing for speech and voice.
“We started coding and hacking together voice experiences and quickly found that we needed to focus on the creative element more than ever before,” he said. “Instead of asking open-ended questions, like, ‘What are you in the mood to eat?” for example, it is important to offer recommendations based on specific options, like, ‘Would you like beef, chicken or vegetarian recipes?”
Here’s a day in the life of Hansen, slightly edited for clarity.
7:15 a.m.: I am up and getting ready for the day with quick breakfast at home, and I ask Alexa for the news flash briefing. I have customized my own sources, including NPR, CNN and some tech blogs.
7:55 a.m.: On the way to the office, I drop off my son, who’s in the second grade.
8:15 a.m.: I arrive at the office and start the morning by catching up over Slack with a few of the project managers in Rain’s New York City office — they are already a few hours ahead of me!
8:45 a.m.: I have a quick meeting with our mobile development team about new support requests regarding the app we built called Reverb. A few users have emailed requests to get weather information for cities aside from their default locations. I discuss potential changes with the team and how we might be able to make the experience easier for users.
9:00 a.m.: I get on a call with Rain’s voice-experience team to discuss requirements for a potential project. We focus on the differences in experience between a chatbot response and voice-only response for the same project. My mind wanders for a second as I get thinking about how we’d minimize the amount of separate code, while still allowing for those differences. On chatbots, for example, we can minimize text and have quick menu prompts. But the voice experience is usually longer, since it’s conversational.
10:00 a.m.: I hop on an executive call to connect with the heads of all our departments on what else is going on at Rain. My portion is, of course, to provide updates on the development department.
10:30 a.m.: We have internal project scrums for our client, Campbell’s. We review everyone’s progress and recap the pending deadlines. We have done a lot of different strategy, creative and development projects for this client — from mobile apps, to websites and an Alexa skill.
Meanwhile, on Slack, I’m alerted to another project where some new requirements will affect a presentation at 4:00 p.m. I jump in the channel, and discuss a few of the options with the team. On Slack, we keep separate channels for each individual project. There are always more than a dozen conversations occurring, so I rely heavily on those @mentions.
11:30 a.m.: I have a meeting with a client to demo the updates that we completed for them over the last two weeks. It depends on the scope and duration of the project, but we typically have two to three demos for each. We have a high-level summary, discuss the goals and then do a live physical demo. We’ve even done some mock demos for clients during new business pitches, and it’s definitely helped us win some clients.
I am excited that the mid-project demo went really well (and we kept the meeting under 30 minutes!). After finishing the client portion of the meeting, we have a quick sprint planning session to review work for the next 2 weeks with the team internally.
12:30 p.m.: I break to find lunch across the street. I head to Blaze and pick up one “Fast-Fire’d” pizza to go and bring it back to the office so that I can catch up on a few emails.
1:00 p.m.: I jump on a conference call with Amazon’s skill certification team to discuss the latest features we completed for the latest Alexa skill we just built. This is usually the last stage of a project, where once we’re done with everything on our end, they vet everything. For this project, the back-and-forth involved deciding whether or not to allow the skill to have a default message in case it did not understand what the user was saying.
2:00 p.m.: I finally have an hour clear on the schedule. I jump into the code of one of the WordPress sites we have nearly finished and make sure the sitemap is updated.
3:30 p.m.: I have a Google Hangout with Rain’s tech lead in Nicaragua to provide a few updates and take a pulse check on how the developers are doing in that office. About half of our development staff is in Nicaragua, but we keep teams mixed between there and Utah. Beyond our constant communication on Slack, this structure is made especially easy because Nicaragua is in the same time zone.
4:00 p.m.: I have a call with one of our clients to review the technical architecture and hosting requirements for an upcoming Alexa skill project. We plan to host the skill using Lambda on AWS, and we need to review security guidelines.
5:00 p.m.: I have a meeting with Rain’s CDO Andrew Howlett to catch up on just about everything. Between his travel and my own, it feels like it has been a month since we were both at the office on the same day. We discuss projects, internal affairs (including plans for our new office building in Nicaragua), and new tech emerging in the news about AI and voice design.
6:00 p.m.: It has been snowing all week, but Alexa says it should be clear tonight. With that, I pick up my wife and son, and head to Sundance Ski Resort for some night skiing, where the resort lights up the main trails till 9:00 p.m. It is only 20 minutes to the resort, and we’ll grab a quick dinner on the way.
9:00 p.m. Our toes are frozen. It’s time to get back in the car and head home.
10:00 p.m.: The family is in bed, and I ask Alexa how my Yahoo Fantasy football matchup is predicted to do on Sunday. I catch up on some more of today’s tech headlines, review new work estimates, and get going on some emails back to the team before going to bed.