Low Latency Natural Language Processing of Corporate Earnings Announcements

This month we have a NLP talk in the financial realm. Andrew Brook and Ashfaq Rahman will be presenting "Low Latency Natural Language Processing of Corporate Earnings Announcements".

Here's the abstract and their bios:

Modern financial markets are highly automated and operate on timescales measured in microseconds - but are significantly impacted by news, regulatory filings and other complex, unstructured information.  This has created a need for systems that can process unstructured data at high speeds and extract key insights with high accuracy.  End-to-end system latency on the order of 10 milliseconds with precision scores in excess of 99% are now possible, even on complex content such as extracting financial metrics from corporate earnings announcements.


Performance at this level requires a holistic approach to the design of the algorithms as well as the overall platform architecture.  In this talk, we present a number of architectural and algorithmic techniques that enable this level of performance and accuracy.  We discuss optimizations that are specific to the financial news problem domain as well as generalizations for broader application.

 

Andrew Brook

Andrew Brook is the Chief Technology Officer of Selerity, responsible for the overall tech strategy at the firm including platform architecture, recruiting the engineering team, and working with clients to understand their needs and defining the platform architecture to meet those requirements.  He was one of the original developers of the Selerity platform, coding the real-time data processing framework and designing Selerity's ultra-low latency messaging middleware.


Andrew has a BS in Computer Science from the University of Illinois and was working on his Master's degree when first bitten by the start-up bug: applying novel optimization and control theories to agile manufacturing systems.  Following the dot-com bust, he moved into finance, building risk management and electronic trading systems for large investment banks.  The lure of the start-up scene eventually became overpowering and in 2009 Andrew joined the newly-formed Selerity.  When not working or geeking out, Andrew enjoys spending time outdoors with his family.

 

Ashfaq Rahman


Ashfaq Rahman is an entrepreneur, computational linguist, and Selerity's Senior Software Engineer / Team Lead for Advanced Algorithms.  Ashfaq defined Selerity's Natural Language Processing and Machine Learning technology and built it from ideas on paper to a market-leading system.  He leads a distributed team of developers in North America and Europe who continually expand and improve the system's capabilities.


Prior to joining Selerity, Ashfaq held a number of prominent positions as both a researcher and an entrepreneur.  Ashfaq was an engineer at Microsoft building server side algorithms, systems and tools. He was the Chief Scientist for the social network lead generation / matching and recommendation engine for shop.com, a computational linguist for Vantage Laboratories building iseek.com search engine, the founding CTO of LoopAnalytics, a recommendation-driven mobile advertising platform and also played a founding role for LocalBlox.com, a hyperlocal social network and content aggregator.  Ashfaq attended graduate school at Columbia University and the University of Pennsylvania with a Masters degree in computer science. He holds a number of software patents.

Join or login to comment.

  • Pankaj T.

    I missed it. Is webinar available of this meetup?

    February 8, 2013

  • David L.

    technical speaker was very slow, making it hard to pay attention

    October 21, 2012

  • Andrew S.

    Interesting talk.

    October 19, 2012

People in this
Meetup are also in:

Imagine having a community behind you

Get started Learn more
Rafaël

We just grab a coffee and speak French. Some people have been coming every week for months... it creates a kind of warmth to the group.

Rafaël, started French Conversation Group

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy