Bring Your Own (Model): Open Adaptability Utilizing spaCy

April is a month that epitomizes Spring; all over there is new growth, change, and new opportunities.  One benefit of having a leadership role in the company, is that I get to name the theme of our releases.  Earlier this year we (I) decided that 2021 would adopt the use of great sporting fish; specifically those often pursued with a fly-rod.

I chose Brook Trout for our May release, as they are an amazing fish that has found a way to adapt to tremendous change.  Brook Trout are native to North America, and survived over-logging, over-fishing, and industrial impacts.  Even when beautiful, native fish like the Grayling were rendered extinct in places like Michigan, Brook Trout adapted even when other non-native trout species were introduced

While we have developed our own NLP; and broad set of linguistic patterns and models in Speech2, covering a range of information risk and insight use cases; we recognize that the ability to learn and adapt with different clients and use cases is critical.  It is nearly impossible for any company, regardless of its size or budget, to possess sufficient clairvoyance; while simultaneously trying to maintain control of all derivative IP.  And because our analytics architecture is open and extensible, it was natural for us to adopt a Bring Your Own (Model) framework.

Thanks to our great engineering team and leadership by our CTO Loren Wheale, we pulled this into the month of April, and is now part of our Speechsolution.  

  • The first thing we did was also incorporate spaCy models as options for NLP and sentiment, and ability to leverage any other supported spaCy models
  • We then extended the framework that allows us to adapt other models that our clients have developed, or are part of industry, government, or academic work

This is all about adapting to a future difficult to predict; but knowing that our clients look to an open, flexible framework as they plan for years to come.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: