T121 Alexa-Aladdin

Voice/Alexa powered financial/stock performance analyst powered by Aladdin API

Comment 2

Categories :

  1. Best Use of the Aladdin API
  2. LA Hacks Challenges a. "What’s one problem you face daily. Solve it at LAHacks" - Analyzing trends/patterns by technical traders. They invest more than 8 hours every day looking at their screens. b. "What’s considered boring by millennial. Exponentially increase engagement for that thing." - Want to become a trader but certainly don't want to gaze at a monitor everyday for the rest of your life. That's why just ask Alexa to do it for you!
  3. Best first time hack (first hackathon for 2 members of our team)

Inspiration

Understanding stocks and identifying patterns is difficult. Even if one is technically savvy, looking at the highly verbose tickers everyday for hours can result in fatigue. Our solution is to create a voice-enabled assistant using Amazon Alexa, that can answer not only general trends/patterns and comparisons but can also answer queries related to prediction using well defined technical indicators (sma, ema, macd, rsi, stochastic etc.)

What it does

Provides a voice-controlled-assistant to the verbose Aladdin API.

  1. It can answer specific queries about stocks' performance over a range of time.
  2. Compare performance of multiple stocks in/over date-ranges.
  3. Analysis and report seasonality of a stock and hence, thereby suggest best and worst months to invest in a stock.
  4. Answer queries related to highest performing stock for a given range of time.
  5. Answer queries related to the best year for a given stock over past 5 years.

How we built it

  1. Nodejs
  2. Amazon Alexa SDK
  3. Firebase
  4. Blackrock Aladdin API
  5. AWS Lambda

Challenges we ran into

  1. Understanding, implementation and integration of algorithms (seasonal, sma etc.)
  2. Lack of domain knowledge and financial jargon required for accuracy verification
  3. Aladdin API instability (reported to Blackrock team :-)) and lack of documentation about type(s) of request(s) & corresponding parameter definitions.
  4. Alexa - parsing user queries into valid date ranges for e.g. last 2 years, last 2 months, in 2014, from 2014 to 2016 etc.
  5. Performance issues when calculating the best stock of all companies to buy/sell.
  6. Amazon alexa dataType issues for recognizing years. It was recognizing the speak/voice for "3 years" as 3016.
  7. Nodejs require is expensive, especially for variable driven requires. Make sure to cache the required data-set.

Accomplishments that we're proud of

  1. Understanding of complex financial indicators which are used for predictions (sma, ema, macd, rsi etc.)
  2. Implementation of trend analysis through seasonal patterns

What we learned

  1. Financial market basics & jargon
  2. Complex indicators and their usage
  3. Aladdin API
  4. Alexa SDK (state oriented version)
  5. Nodejs

What's next for T121 Alexa-Aladdin

  1. Integrate the remaining 3 APIs from Aladdin (Portfolio Analysis, Securities & Data)
  2. Comparison of a stock with its' corresponding ETF
  3. Better prediction using more advanced analysis indicators

Built With

Share this project:

Updates