Past Meetups

We meet about once a month to share how we create ML + UX products. Check out a few of our past events below.

 

SAP.io + Noodle.ai: Navigating the intersection between Product Design and Machine Learning

Tony Chu – Principal Designer at Noodle.ai

With all the hype around ML-driven this, and AI-driven that, one might mistake machine learning for magic you simply sprinkle on products. To truly take advantage of machine learning, however, we must understand how the strengths and weaknesses of ML, and how that maps to product experiences. In this talk, we will look at how machine learning creates new opportunities and new problems in product design, and look at some examples of ML applied in products we use today.

 

Panel Q&A: Data Science in Product Development

Elizabeth Brook – Product @ Noodle.ai, Jenn Gamble – Data Science @ Noodle.ai, Tony Chu - Design @ Noodle.ai
Moderator – Nikki Helmer – Data Science @ SAP.iO

Following the presentation, a cross-functional panel from Noodle.ai will talk about bringing business thinking, data science and design together in a product development process. They will answer questions about how they frame problems and arrive at solutions together, and about the work that Noodle.ai is doing to solve problems for large enterprises.

 

Clara Labs Tech Talks: Human-in-the-loop machine learning

Algorithm aversion in human-in-the-loop automation systems
Jason Laska, Ph.D. -- Engineering @ Clara

Clara Labs is an email-based scheduling service for busy people. Simply CC Clara on an email to a person you want to meet with, and we'll handle the back and forth game of email-tag for you in accordance with your preferences. To build a robust and accurate system that gracefully handles nuanced requests, we've combined machine learning (ML) with a distributed human labor force.

In this discussion, we'll review recent results in the literature on algorithm aversion: the idea that people reject algorithmic solutions to problems even when knowing such algorithms are more performant on average. Next, we'll explore a case study of this effect in Clara's scheduling system with regards to our automation system for suggesting meeting times.


Interfacing with human-in-the-loop systems
Gavin Schulz -- Engineering @ Clara

Traditional scheduling assistants (whether remote/virtual or in-house) are a single person. When that person joins your organization you spend several weeks getting them up to speed on your personal meeting habits. This often happens organically through your interactions and feedback and without significant thought about your scheduling process up front. Unlike a traditional assistant, with Clara you configure your assistant via software preferences at any time, including immediately when you join the service. This is more akin to an app than a person.

In this discussion, we'll talk about some of the challenges with building a preference configuration interface and on-boarding customers for a conversational agent. Topics include: determining which scheduling habits should be preferences, explaining to customers the importance of engaging with these customizations, and messaging how one should expect the product to behave depending on the configuration.