Case Study: Carbon09 Voice Trading for Amazon Alexa

Brian McCallion Case Study, Internet of Things 0 Comments

About

Carbon09, Inc. is a software firm focused on bringing next generation technology to financial markets. Tradier is a brokerage API provider dedicated to delivering choice, value, and innovation to the self-directed financial market through open and simple technology to foster a vibrant community of traders, providers, and developers. Bronze Drum is a New York based consulting firm focused on Internet of Things, real-time applications, real-time analytics, and working with financial services customers to leverage these technologies.

Challenge

While professional investors benefit from access to sophisticated tools, the retail investor may find simple hedging strategies time consuming to plan and execute. How many times have you been fortunate enough to have a bet on a stock pay-off, and then either sell it early to lock-in wins, or sold it late, after seeing the price drop significantly. As humans we’re wired for fight or flight. The literature provides advice, such as “buy and hold” and think long term. Yet for many investors hedging equities with options is a complex task. What if this task could be made easier? What if it could be done while unwinding on the couch after a hard day’s work? After a full day at the office many of us return home to a brief respite, and then return again. In the interim we prize our leisure and family time. To demonstrate how voice interfaces can enable new types of users to augment their own skills with those of a computer, Bronze Drum asked Tradier to provide access to the APIs. Tradier’s Dan Raju, responded enthusiastically, and provided Bronze Drum with access to the APIs and a Production Tradier account, production-like in every way except that orders don’t actually reach the exchange.

Solution

To maximize the time and focus on the application, Bronze Drum deployed a Continuous Delivery pipeline using CodePipeline, Jenkins, and Github. In this way, the Alexa Voice application is packaged, then deployed with a simple code check-in.
The application enables Amazon Echo, Alexa users to place trades by voice, to get quotes, and to request news. While there are many ways to access quant functions, the team decided to use AWS Lambda Node.js functions transcompiled from the open source C++ library, QuantLib so as to introduce these functions to a broader audience. The functions enable the same code to execute in the browser, in a Lambda function, and in a mobile application environment.
The simple hedging algorithm looks at existing positions, and if unhedged, selects options according to the user’s tolerance for risk and the cost of the “insurance”. Users can create and manage watch lists, and access their Tradier account investments by voice, and buy and sell equities via the Alexa Voice interface. The application persists the application state in DynamoDB, using the updated Alexa SDK.

As many of the api calls retrieved rely upon the same set of relatively slow changing data. Amazon API Gateway enables Bronze Drum developers to create very specific API calls, and to pass in parameters specific to the application requirements, and simplify the client-side code.

Results and Benefits

The Tradier API enables developers to build incredible fintech applications without requiring the developers to spend substantial capital. Today developers can leverage the Alexa voice api to navigate the positions in their portfolio, evaluate options, and purchase insurance against downside risk. The experience reduces the concentration required to perform such tasks and navigate multiple screens to a simple voice request. By leveraging Alexa, and the Tradier API, and Amazon Web Services, we built a simple yet sophisticated Alexa skill to assist the retail investor.