Thursday, January 10, 2019

"A Bloomberg Terminal Alternative That Uses AI"

From Nanalyze:
Going to business school is a great time to “pivot” your present career into something more exciting and more lucrative. Among the target jobs that freshly minted MBAs pursue, careers in finance top the list when it comes to something that will impress your family and get you
laiddates. Business schools often have a department called “career services” which is supposed to help prepare you for your future career in finance. Practicing interview skills, networking with alumni, and learning about the tools of the trade, are all great ways to prepare yourself for your 10-year stint in the big leagues before you get burned out and go off to “do something on your own” – which usually means you then turn right when you get on a plane instead of left. One tool that you’ll inevitably come across during your illustrious career in finance is the insufferable Bloomberg terminal. If you don’t know what that is, we’ll help enlighten you.

An Intro to Bloomberg Terminals
Imagine if you will, that you took some completely archaic technology – like Windows 3.1 – and then created an entire financial data platform around it, along with a custom keyboard that required a complex set of key presses to do even the simplest thing. In order to design the keyboard, imagine that you hired a bunch of colorblind Bangladeshi high school kids with only a basic command of the English language and asked them to create the user interface (UI) from scratch in 24 hours, making sure to include lots of cryptic color-coded keys. The end result would be a UI that even people with some of the highest IQs on the planet would struggle to use. Once that’s been accomplished, you can then move on to screwing up all visual aspects of your data platform next.

Beautiful, crisp, clean, high-definition graphics? Absolutely not. Go with 256-colors max, and design the whole thing using ASCI characters with lots of annoying text highlights like this and text colors like this and then suddenly just start changing the colors as you go along for absolutely no reason at all. Make sure it’s as painful to look at as it is to use. The end result would look something like a Bloomberg terminal does today. Here’s what one looks like in all its glory:
 Fugly as sin and as intuitive as the female brain – Source: Brown University
[The "intuitive" bit is sadly mistaken, most women can intuit faster than most men, I don't know why}

Ask anyone who uses one of these things if they enjoy using it and people might say they do but inside they’re seething with rage every time they place their hands on that lego-looking keyboard. As David Friedman says on a Quora response:
  • The Bloomberg Terminal interface looks terrible, but it allows traders and other users to pretend you need to be experienced and knowledgeable to use it.
Exactly. Of course, once you get used to using it – which only happens after you digest the 86-page manual and practice for weeks on end – “you’ll never look back.” But what if we took this horrible contraption that hasn’t been redesigned since Michael Jackson was beating it back in the early ’80s and completely sidelined it in favor of some other solution? Perhaps, we might even consider using some of today’s advanced technologies – like machine learning – to create a Bloomberg terminal alternative that was actually usable and allowed someone to perform at higher levels with a minimal learning curve. That’s what a company called Sentieo claims to be doing.

Sentieo – A Bloomberg Terminal Alternative

Founded in 2012, San Francisco startup Sentieo has taken in $25 million in funding so far to develop a “modern financial research platform” that’s powered by artificial intelligence. Founded by former hedge fund analysts, the tool is already being used by over 700 buy-side, corporate, and sell-side clients, and utilizes machine learning and natural language processing to “instantly surface textual and financial insights.” Their product launched in 2015, and just a few months ago they closed the lion’s share of their funding in the form of a $19 million Series A round.
The finance industry is often slow to adopt new technologies, and there are firms out there today with teams of analysts that to do nothing but pore over financial filings (10Qs, 10Ks, and the like) searching for certain bits of data. For example, they may be tasked with looking at a firm’s revenue streams so that they can classify it into a particular industry. These types of abstract data points can now be retrieved using Natural Language Processing (NLP). They’ve also expanded beyond traditional sources of data and opened up the platform to alternative data sets that can be used to generate alpha:
By tracking, scrubbing, and visualizing key company/brand data such as website traffic, search volumes, Twitter mentions, app downloads and more, Mosaic allows analysts to correlate web interest to revenue growth and price performance, driving a significant alpha generation edge
They’re not just using NLP for screen-scraping, but also enabling users with the ability to ask the platform questions using natural language:...
...When it comes to pricing, Sentieo claims to be “aggressively priced” versus their competitors. An article by TechCrunch around the time of their last funding round put pricing at $500 to $1,000 per month, per seat. That means they’re somewhere around half of Bloomberg’s “$21,000 per seat” annual price tag. They also need to compete with other large financial data providers in addition to Bloomberg. Thomson Reuters Eikon is said to be the closest competitor to Bloomberg...

I don't see any mention of or Goldman backed Symphony.
For those you may wish to peruse:
October 2018 
"Twilight Of The Terminal: The Disruption Of Bloomberg L.P."
November 2017
Stalking Bloomberg: " now has more than 80,000 subscribers"
July 2017 
"Can Anyone Bury Bloomberg?"
October 2016 Hires Former Bloomberg Honcho to Take On Bloomberg Terminals With Machine-Generated News Bulletins and Stories
Aug. 2014
A Deep Dive Into Goldman's Chat Platform (GS)
March 2014 
Can the Bloomberg Terminal Be Toppled?
February 2014 
Media/IT: The Race to Topple Bloomberg