Using ROSS for Florida Lawyers

Charlie von Simson
June 27, 2019
Legal Research

The entire ROSS team is excited that our artificial intelligence legal research platform has become the newest member benefit of the Florida Bar. Several ROSS attorneys are attending the Florida Bar meeting beginning today. We hope we’ll have the chance to meet many Florida lawyers and introduce them to ROSS.

Even though Florida lawyers will be able to get up and running immediately, our team of litigators thought that a Florida-specific example of how ROSS works might help everyone get started.

This post will barely scratch the surface by introducing new users to the ROSS natural language search capabilities with an example drawn from an issue that has been actively litigated in Florida and is now awaiting a decision from the Eleventh Circuit.

Labor and employment lawyers across the U.S. have been watching the developments in Gil v. Winn-Dixie Stores, Inc., 247 F.Supp.3d 1340 (S.D. Fla. 2017), one of the few ADA website accessibility cases to go to trial. The district court ruled in favor of Gil, holding that Winn-Dixie’s website was a place of public accommodation inaccessible to persons with visual impairments and held that: a.) the website had to be fully accessible; and b.) Winn-Dixie was required to implement the Web Content Accessibility Guidelines (WCAG) 2.0 for its website by an agreed-upon date.

The issues raised by Gil present an interesting research assignment for ROSS.

We start with the query: When is a website a “public accommodation” under the ADA in Florida since 2015?

You’ll notice a few things about ROSS natural language search that sets it apart from traditional platforms:

  • ROSS understands natural language queries drafted the way you would talk to a colleague. That’s the underlying power of artificial intelligence. When framing your queries, be as specific as possible. Queries between 14 and 25 words work best.
  • The ROSS named entity recognition filters automatically understand that you are looking for cases from Florida that have been decided since 2015. You don’t need to click through rows and rows of check-box filters to get to the correct jurisdiction and date range. 
  • You can use quotation marks as part of a natural language search to emphasize a particular phrase and ensure that every result will return the phrase as you’ve identified it. Quotation marks can be helpful in narrowing searches and detecting terms of art.
  • ROSS allows you to search by the procedural posture of your case by entering up to four different motions. Here we’ve entered motion for summary judgment and motion to dismiss simply because they tend to be the most commonly searched. ROSS automatically recognizes all motions that are filing events in the federal electronic filing system. 
  • ROSS lets you enter fact patterns of up to twenty words. The ROSS algorithms will consider that fact pattern in selecting and ranking the most relevant cases. The fact search feature in ROSS addresses the difficulty lawyers often face when looking for cases decided on specific facts. Here we’ve entered “screen reader software with a keyboard” because many of the cases involve such software and many visually impaired people are unable to effectively use a mouse. 
  • You can also select terms to exclude from cases responding to your query. The exclusion filter is particularly useful when a word or phrase designates an element of a case that you know is irrelevant to your issue. Here we’ve eliminated “consent decree” because cases decided based on the terms of existing consent decrees are often not decided in the same way as cases before court de novo.  

The results from this query are interesting. We’ll just summarize them here.


The first result ROSS returns is an important website accessibility case decided under the ADA. In Haynes v. Dunkin’ Donuts, LLC et al., Case No. 18-10373 (11th Cir. July 31, 2018), the Eleventh Circuit reversed the lower court’s dismissal of the website accessibility complaint of Dennis Haynes, a legally blind person. The lower court had concluded that Haynes had failed to properly allege a nexus between barriers to accessing the website and an inability to access services and goods at a physical store. In reversing and remanding, the appellate court found that Haynes had demonstrated “a plausible claim for relief under the ADA.”

ROSS clearly identifies the elements of the case that made it relevant to our query. The title and citation to the case appear at the top left of the result. Although the case was closely followed, Haynes is an unpublished opinion. If you’re only looking for precedential authority, you can filter results to a publication status. Find the filter in the drop-down menu under “Publication Status.”

The bars at the top of the case indicate that the decision is a “Deep Match”. That means the ROSS algorithms have a high degree of confidence that the case is relevant to the query. The Motion Match indicates that the case was decided in the same procedural posture as one of the motions you’ve entered. The Fact Match lets you know that the case involves screen reader software and use of a keyboard.

Of course, the Gil district court case is among the top matches.

Although the primary elements discussed above are included in this result, you’ll note that Gil is not a motion match. That’s because it was decided as a bench verdict, not on motion. The ROSS algorithms correctly detected the posture and ranked it a few cases below Haynes because it is not a procedural match.

As we mentioned, this quick summary barely scratches the surface of all the features that will help Florida lawyers be more efficient advocates and add value to their client relationships. If you stop and see us during the conference we’ll be happy to show you everything else ROSS has to offer. You can also click here to start your 14 day free trial of ROSS.  

Charlie von Simson

Charlie von Simson is a legal subject matter expert at ROSS. He practiced law for twenty years before running away to join a startup.