Small Data and the Model Stack

It's never been easier to wrangle and extract value from huge data… but how should a startup think about getting value from data streams that are just getting started, or find insights about small partitions of data for individual customers?I'll present on how we've tackled these challenges at frame.ai, where we're focused on providing realtime insights about realtime chat streams within one company at a time. I'll talk about setting metrics focused on time-to-value for customers, leveraging model stacks trained on larger data sets, and adopting NLP analysis techniques that behave well where data is sparse.