Cloud9 Technologies Cracks the Code on Trader Voice Transcription
July 27, 2017
Cloud9 Develops Groundbreaking Solution with Quantiphi and Google Open Source Technology that Provides Insightful Voice Data
Cloud9 Technologies, the leading voice collaboration and analytics platform for financial firms and institutional traders, today unveiled the industry’s first purpose-built voice trading transcription solution. The Cloud9 Voice Transcription Solution, developed with cutting-edge deep learning technology in collaboration with Quantiphi and utilizing Google Open Source technology, provides firms with the ability to harness the insight in their voice trading recordings and data by converting speech to text. The groundbreaking solution will revolutionize compliance, surveillance, and advanced analytics in the trader voice space.
“Today, PMs, traders, and analysts utilize only 5% to 10% of voice data for analysis – with Cloud9’s Voice Transcription Solution, they can now use 100%,” said German Soto Sanchez, Global Head of Corporate Development at Cloud9.
Cloud9 solved the complex problem of trader voice transcription by collaborating with Quantiphi and leveraging Google’s TensorFlow technology to build a custom model that can detect and transcribe trader voice conversations. Historically, this process has proven impossible to address using conventional transcription tools, with existing industry offerings falling significantly short of their stated accuracy and performance.
“Trader voice comes with a unique set of problems for transcription – it’s full of slang and jargon, it includes overlapping speech and pitch variations, and the audio clips tend to be only a few seconds – all which make it incredibly difficult to transcribe. Typical transcription models and approaches do not work for the financial services industry,” said Asif Hasan, Co-Founder of Quantiphi. “Cloud9 had to build its model from the ground-up using advanced, deep learning technologies. The result is, frankly, impressive – we are able to transcribe multiple trader voices at unprecedented levels of accuracy within milliseconds.”
This custom-transcription model can be applied to other highly-specialized, vernacular-heavy areas like air traffic control, legal, and healthcare.
Cloud9’s transcription solution debuted at a live demonstration in front of a capacity crowd composed of representatives from major financial services firms, technologists, and industry consultants. The Cloud9 Voice Transcription solution represents the Company’s first of many analytics offerings that will leverage its expertise in deep learning.
“Cloud9 Trader Voice Transcription will be game-changing for this industry,” said Richard Bryant, a fixed income trader and capital markets executive with over 20 years of trading experience. “To hear an actual trader voice call and then to see the model identify the complex jargon and precisely transcribe the conversation is truly remarkable. Up until now, since many had promised and failed to provide such a solution, I did not think this was remotely feasible.”
“Legacy voice turret infrastructure simply cannot support transcription at this standard,” said Soto Sanchez. “The Company’s highly advanced cloud based platform represents a quantum leap in voice trading technology and was critical in providing the unique inputs necessary for the creation of this sophisticated solution. With our transcription solution, we’ve unlocked the potential within voice to power analytics and help firms develop greater insight into their trades and trade activity.”
“In short, the future of advanced trading is through the power of voice, and by working with Google technology and the top data science experts at Quantiphi, Cloud9 will be the driving force of this industry.”
Check out highlights from our machine learning meetup, where we debuted the Cloud9 Voice Transcription Solution.