The World Famous AI Smart Coffee at CLEUR 19

So we took the now “World Famous” (possible exaggeration) AI Smart Coffee app to CiscoLive Barcelona and it was a blast!

Using a combination of Cisco Webex Teams, Google Cloud Vision, Kubernetes ,IFTT and some creativity, we analysed 837 faces, served 837 custom cups of coffee, and met 837 happy customers.

CLEUR Fatigue and Caffeine

Screenshot 2019-02-03 at 09.03.13There was a great buzz around the booth and our team of moderately trained “baristas” kept crowds entertained, educated and caffeinated. We even wore aprons!

We saw science at work! Average fatigue levels rose from the morning to afternoon, and there was a marked rise from Monday to Thursday after four days of learning and four evenings of entertainment. People came back multiple times to see how their faces were faring during the week. Some tried to trick the system with smiles, frowns, lighting and makeup.

We had a an unlikely team of baristas running the booth. Consulting Systems Engineer Tony certainly found his calling pouring coffee for the masses and delighting punters with his East End banter. NetDevopsguy Simon was snapping and brewing between running DevNet labs, taking exams and having a vape. Pedro layered on the latin charm whilst being both tech lead and cleaner. I would say I was responsible for customer experience and avoiding any form of manual labour.

The Dream Team: Roger, Sara, Tony, Simon and Pedro

Screenshot 2019-02-03 at 09.17.34.png

Sara, our most excellent assistant for the week, went from badge scanning duties on day one to explaining the intricacies of Machine Learning and application flows on day two. This shows how a simple fun demo can be a powerful tool for education, demonstration of application architectures and the art of the possible.

A Bit of Creativity and Design Thinking 

Each cup of coffee took around 2 minutes from Face to Cup. 15 seconds for analysis and 2 minutes to brew. (oh..we didn’t skimp on the coffee, we used good quality ground coffee my friend !!). Brewing time gave us a great opportunity for discussion, and what was so amazing is just how many customers shared ideas, use cases, applications and integrations for the solution. Some of these ideas we may integrate in the next iteration of the demo and some we are taking to a Cisco Business Unit for a potential product integration. (the emotion and fatigue Machine Learning elements, not coffee making. I don’t think Cisco will be moving into the coffee business just yet)

The Application Architecture

Screenshot 2019-02-03 at 10.28.23

Our coffee demo is a great example of a modern microservices application architecture. Distributed, leveraging many disparate API driven services from on-premise to cloud.

Next destination is DEVNET Create in April. We’ve also been invited to Russia, New York and San Francisco. However, the weak link of this application is the Human API. It’s totally exhausting work and our fatigue levels by the end of the week broke the ML model. Maybe in 2.0 we will remove the human element, but that being said, how could we possibly replace the legend that is Tony the East End barista (Starbucks now has his resume’)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s