Alternative Data Newsletter #96 – November 29th, 2018

Announcements at our conference in NYC today


What Datasets Are Getting Traction?

  • Quant: this vendor provides Artificial-Intelligence-as-a-Service for investment analysis and trading. The company creates custom portfolios for clients by combining fundamental analysis, quantitative metrics, sellside reports, macro news, and money flows. Our Data Sourcing clients can view the full profile here.
  • Discretionary: this vendor delivers credit/debit card data covering a panel of 15 million people across 50 US states. The company offers raw feeds and dashboards. Our Data Sourcing clients can view the full profile here.
  • New: this company extracts event level semantics from public news. Their dataset is on entity plus event level and covers over 15,000 global stocks, all currencies and major commodities. Our Data Sourcing clients can view the full profile here.

Data Science Lab

  • Open sourcingnbformat – this lesser known part of the Jupyter notebook ecosystem allows you to programmatically create Jupyter notebooks. Have a style convention or standard content you want to present in the Jupyter notebook medium? This library allows you to build a notebook with code and markdown cells using a simple python API.
  • What we’re reading: Training Sequence-to-Sequence Models to generate Github issue titles – this in-depth tutorial provides code and a full explanation of the steps involved in building a sequence-to-sequence encoder-decoder model to ingest Github issue body text and output a summarized title. The article goes through how to get the Github issue dataset, cleaning the text data, training the model, and a sample of results using the model. It also touches on alternative uses for the model such as finding the most similar issues to a given issue. Useful applications in the financial industry would be document summarization and topic inference in news or other natural language sources.

Legal & Compliance, Efficiency Improvements & Best Practice

  • Efficiency Improvements – Standardized Metadata: contact to get access to the first document standardizing the metadata schema.
  • Best Practice – Survey: in November we are surveying clients on compensation packages for the different roles associated with alternative data. The results of the survey will be published on a redacted basis following the Data Forum event on December 6th. To learn more contact

Updates For Alternative Data Vendors

  • Increase revenue: we are launching several new features to help alternative data vendors succeed. To learn more email our CEO –
  • DDQs: we have started a working group to standardize dataset due diligence questionnaires. If your firm would like to participate email
  • Dashboards: to help you access traditional fundamental investors we are building customized dashboards on top of alternative datasets. Contact us to learn more at
  • Funding: if your firm is seeking to raise capital contact us – we can help. Email:

Notable News in the Alternative Data Space