This is a prototype, it has.....

- One asset class (stocks)

-5k tickers

- Low keyword relevancy

- One anomaly detection strategy

- No predictive models

- No quality checks (filtering, etc)

No visualisation, analytic tools, APIs, etc

...while the end product will ....

- Cover several assets clases in the anomaly scouting/prediction scope

• • Liquid; equities, bonds, macrotrackers, commodities, currencies, non-US exchanges
  • Illiquid; Real Estate, Private Equity/Credit, Startups and Early Stage
  • Non-financial and alternative data (political, social, technological data that may catalyse opportunities)
  • Curated Sentiments
  • Proprietary data - otherwise non existent data (eg- specialized niches, or geo markets) that the platform gives access too

- Natural language inputs with high relevancy

- Tens of different anomaly detection/prediction strategies that can check in n dimensional spaces made of a mix from all data sets above according to an customizable critera

- Prediction Model where an AI will help recommend data sets based on the NL input as well as learn from the past correlations to predict a move or an anomaly

- Quality checks, noise filtering and other filtering on the data sets.

- A large library of analytical tools and visualization that help the investment manager's decision making process.