in

The Continued Rise of Speech Recognition

Mike Kelly

There is a resurgence of interest in speech recognition (SR) within the legal profession currently. This same interest incidentally is mirrored in other professions such as healthcare and in business generally.

It is a trend that is likely to continue. Refer to the 2014 Gartner report where Speech Recognition is now positioned as one of the leading productivity tools of our age.

Penetration

SR is now penetrated far and wide within the legal sector. There are installations of all sizes…many sole practitioners and a number of firms with just one or two users. There are also an increasing number of firms with 5-10 licences and a few with 20 plus. It is difficult to be 100% certain but we estimate that there are close to 1,000 SR licences in the field today. The product is extensively used with email but also with precedents and templates and of course for substantial opinion or advice related matters.

Motivations

The motivations are varied. In many cases it is driven by individual lawyers or their Practice Managers…interestingly not too often by IT at this stage, but this is set to change in my view now that the technology is now main stream. My expectation is that we will progressively see the “institutionalising” of SR in firms by including it in staff indoctrination processes and making it either a compulsory tool or one that is available as part of the firm’s standard IT platform.

With some firms today there is no planned approach. A lawyer might just have read about it and want to try…referrals are an increasing driver. Others may be forced into it because of
secretarial constraints or because while they might be used to self-creating documents they find the process of typing slow or tedious by comparison to what speech alone can deliver. Remember you can speak at least three times faster than you can type. In others it is a cold hearted ROI that drives SR.

SR is overwhelmingly used in what is called “front end” mode where a lawyer dictates and the transcribed text appears on their own PC. Initially the thinking was that “back end” deployment of SR, where the lawyer dictated but the text appears on the typist PC, would also be on interest but this has not taken off and in my opinion is unlikely to do so. With SR technology so accurate nowadays, secretarial involvement and the attendant waiting time involved, is a luxury few can afford.

SR is a productivity and efficiency tool

SR is an interesting and clever technology, but first and foremost it is a productivity and efficiency tool. It is not without cost when one factors in installation/training and support plus perhaps a wireless input device, but it will quickly pay for itself if deployed correctly. In one firm where we piloted just the one licence, it quickly grew to nine users and the annual savings of $170k in annual staff support costs is not insignificant. For a one off investment of just $12k this equates to a payback of weeks rather than months.

Because lawyers can produce our own documents on the spot, this can free up typists and secretaries. With the need for typing now diminished, this resource can be redirected into more productive areas as true personal assistants and authors in their own right.

Current scene

To my mind the traditional dictation/transcription market has plateaued. This applies to both desktop digital and network digital technologies. It is not something I like to admit because we enjoy a sizeable business in both areas, but that is my definite opinion. Lawyers no longer want to dictate, send a draft through to their secretary PA and then have it returned for final sign off. They will either type up an original themselves relying heavily on smart precedents or they will do the same things except they will use voice rather than the keyboard. With the advent of a specifically tailored NZ Legal SR software that is now so amazingly accurate there will be few mistakes and the old myth that lawyers won’t format and edit themselves is simply not true. The modern lawyer is quite adept at editing and formatting and as you’d expect in this age of instant fulfilment, get frustrated at what they see as the unneccessary and time consuming step of sending dictated work to a secretary, when they can start and complete a matter themselves in one session.

SR is used for the creation of standard forms type documents where it integrates with a firms precedent system, for the creation of an accurate first draft and invariably with the daily deluge of emails which are fast becoming bread and butter to practitioners.

We are seeing SR used by authors of all ages. While it is probably true that for many more established lawyers the preference is to continue to use a secretary/PA I am continually surprised at the cross section of users out there. Technology or age tends not to be a barrier…but there are other barriers of which a potential user needs to be aware.

Pitfalls

There are a few to be aware of:

 

  • Not all users or all work types suit SR. While age or technical knowhow is largely irrelevant there must be a certain enthusiasm to explore a more productive alternatives and accept a small amount of training. It is not sensible to consider a firm wide roll out without total buy in from all users.

 

  • While dictation speed, volume, background noise etc are not issues, the user must be able to dictate or compose their thoughts on the fly. In days gone by this skill was taught to most new graduates but nowadays it is not that common. However it’s not a bad skill to have for any lawyer, I think you’d agree, and like all skills it can be learned. If you are not adept at dictation, we recommend daily practice as a way of coming up to speed. Here are some tips on dictation.

 

  • The technology is quick to adapt to your voice and produce a highly accurate document but it does need training to extract most value. Don’t be tempted to scrimp in this area. Remember you are spending money to make money.

 

  • If you operate in open plan are you comfortable in speaking aloud in front of your colleagues? Keyboarding a document or email by comparison is totally private by and you are free to make mistakes or correct without anyone knowing. Do remember though that you can dictate quietly.

 

  • Proof reading your own dictated document can be a challenge for some. They find it far easier to offload this task to a qualified legal secretary.

 

New features and direction

Deep learning

Nuance who are one of the leaders in the SR field, designed their speech engine to work off of the principal of deep learning, which relates to artificial intelligence. What this means in practical terms is that it is more tolerant of accents, background noise and adapts your profile on the fly.

Virtual environments

Support for Citrix virtual environments and also TS. Fully Win 10 Office 2016/365 compatible.

Reporting

 

SR is now biting directly into the space occupied by the network dictation solutions employed by many large firms in that for multiple users it is now possible for IT to centrally administer,manage and maintain (modify, repair, upgrade, remove) the SR software licences and track employee usage, redistribute licenses based on employee use and manage or share customisations, including custom words, commands and auto-texts, across multiple SR users.

 

Look for a SR product that allows usage monitoring. Especially in a larger deployment this allows conclusions to be drawn in terms of how effectively SR is being uses. Great for anyone interested in the ROI of the SR investment and for highlighting training issues ets.

 

In Summary

The takeaway from all this is that no longer do users have to worry about:

 

  • Accuracy (it’s quickly up to 95% plus immediately after installation)

 

  • Enrolment. Just takes 2 -3 minutes to adapt to your voice

 

  • Ease of use. Now very intuitive and able to use mousing as well as voice commands.

 

  • Computing power. Any modern computer will have with an I5 processor and 4GB RAM or better will do a fine job.

 

  • SR software can be loaded onto as many as four (4) separate devices meaning you can also store your profile in one place and have it synchronised between your office computer and your laptop.

 

  • With multiple users a firm can measure usage and identify competence and training needs.

 

  • SR can be trialed, so a leap of faith is not necessary.

For opinion oriented first drafts, general commercial law involving precedent and templates and for dealing with the mass of email traffic, SR is worth considering.

Author

mkellyMike Kelly started his business career working for Fletcher Challenge in the 80/90’s  and has been co-owner of Sound Business Systems (SBS) based in Auckland NZ ever since.  Sound Business Systems is a provider of Phillips, Olympus, Winscribe and Dragon SR technologies.  He spends his day to day working with law firms of all sizes on dictation and speech recognition related solutions.

The firm’s customer base is NZ wide and they deal with everyone from the sole practitioner right through to the largest firms in the country. They have a full team of specialists who install train and provide after sales support.

this article was written for  ALPMA in New Zealand.

 

 

50 Cent’s Big Bucks Big Law Lawsuit

Blank Rome Welcomes Corporate & Securities Partner James Barnes in Pittsburgh