GreenKey Scribe: Alexa-like Speech Recognition for Police, Fire, and Emergency Services

Tejas Shastry

April 30, 2018

Over the last couple of years at GreenKey, I’ve been part of the team building Scribe, a deep-learning AI for nuanced industries. Dubbed the “Alexa for Wall Street” by Forbes, we’re now expanding Scribe’s focus to emergency services, like police and fire departments, and emergency medical technicians.

Scribe allows brokers, traders, analysts, and others at financial institutions to transcribe their telephone conversations in real-time and extract important data like quotes and trades with high accuracy. Speech recognition helps financial firms save time, and therefore, save money.

But that’s the tip of the iceberg when it comes to Scribe’s capabilities. There are several parallels that make Scribe work well for finance and public safety:

  • Both environments are noisy.
  • Both have their own “tribal lexicon” of acronyms and keywords.
  • The majority of conversations happen over a telephone – a stream of audio that devices like Alexa can’t access.

Have a listen to the two samples below to see what we mean:

Thank you. VIN number 16646. 16646. 13 Hastings all the way down all blackout. I’ve got no city street lights. Alright 10-4 give me one moment let me take a look.

We’re taking on a risk in this one. We need to buy the UK one and sell him the Aussie with the currency, there’s a currency, there’s a currency risk, there’s a currency involved in it there’s a spread on it as well. The currency we opened yesterday was trading at 16% the DOC was trading at 16%. Right, Jase.


We built GreenKey Scribe specifically to address these issues – transcribing audio directly from telephone voice streams in noisy environments, and enabling firms to teach it the way they speak with ease. Financial traders are constantly listening to complex financial quotes coming off of a turret phone on a noisy trading floor. Police audio faces the same issues of noisy environments over low-quality telephony networks with specific, critical pieces of information (license plates, addresses, police codes).

Focusing on emergency services plays to our team’s strengths, but also comes with challenges. The tight security and compliance protocols for personal identification that we’ve navigated in the financial industry helps us when talking to first responder departments who have the same concerns. However, the level of standards that the financial industry has in their language does not extend to emergency services, where police departments can often have varying definitions for the same police codes. The potential for impact in a life-critical industry like emergency services industry is a reason many of us joined GreenKey, and that motivation has helped us navigate many of these challenges.

The financial industry is embracing Scribe with open arms. Now we’re bringing the same solution to leaders in the emergency service industry by bringing Scribe to police and first responders around the nation. GreenKey helps them react faster to situations by transcribing important information onto the screen in real-time.

Financial professionals have already saved more money using Scribe, and now first responders may be able to save more lives.

Blog

August 2019 Newsletter

Tejas Shastry, our Chief Data Scientist, recently participated in a highly successful webinar with Greenwich Associates where he discussed artificial intelligence and NLP in trading, including his thoughts on potential future uses in structuring data.
Read More
August 2019 Newsletter
Blog

Artificial Intelligence on the Trading Desk Webinar: A Summary

GreenKey’s Chief Data Scientist, Tejas Shastry, was a featured panelist in a July 25 webinar hosted by Greenwich Associates that discussed various aspects of artificial intelligence now being used on trading desks.
Read More
Artificial Intelligence on the Trading Desk Webinar: A Summary
Blog

There are trade signals within your audio data

Identification and interpretation of pace, volume, pitch, and cadence are part of the human brain’s processing ability to identify sentiment, based on years of training.
Read More
There are trade signals within your audio data
Blog

July 2019 Newsletter

This month we highlight our ability to unlock conversations across voice and chat - leading to fewer missed trading opportunities across clients and asset classes - as well as new analytics reports specifically tailored to show desk heads what's trending.
Read More
July 2019 Newsletter
{"slides_column":1,"slides_scroll":1,"dots":"true","arrows":"true","autoplay":"true","autoplay_interval":2000,"speed":300}

By signing up, you agree to the Terms of Use, the Privacy Policy and the transfer of your personal data from your country of residence to the United States (if different).

We promise we don’t send spam.