Published Jul 6, 2021

SDS 485: Financial Data Engineering — with Doug Eisenstein

Doug Eisenstein, founder of Advanti, joins Jon Krohn to delve into financial data engineering, tackling challenges like entity resolution and the crucial role of temporality in trading strategies, while reflecting on his evolution from consultancy to solution-driven product development.
Episode Highlights
Super Data Science: ML & AI Podcast with Jon Krohn logo

Popular Clips

Episode Highlights

  • Resolution Issues

    Entity resolution in the financial sector is fraught with complexities due to diverse data sources and identifiers. highlights the challenge of matching identifiers across various financial instruments like indices, ETFs, and futures contracts, which provide unique metrics but require connection for comprehensive data coverage 1. This issue becomes particularly pronounced as companies transition from fundamental to systematic, data-driven trading, necessitating robust entity extraction and resolution processes 2. Doug emphasizes the importance of capturing descriptive characteristics like name, country, and currency to effectively link entities over time 1.

    We've seen this as a huge problem. And that's one of the areas that we find is the most challenging for customers to really get right.

    ---

    The ability to resolve these entities is crucial for investment managers dealing with multi-asset classes, as they must navigate data from various sources and adapt to changes like mergers and acquisitions 2.

       

    Extraction Insights

    Entity extraction involves identifying and linking specific entities within financial data sets, a process crucial for making informed investment decisions. explains that this involves scanning documents like SEC filings to extract important nouns, such as company names or financial instruments, and linking them using structures like knowledge graphs 3. Doug discusses products developed at Aristos, such as Finflow and Dominus, which address these challenges by managing entity data and linking disparate data sources 4.

    It's the way you describe it is, right? Like, you have to be. You're searching through documents, through structured, through structured data, semi structured, unstructured data.

    ---

    These tools help streamline data flow and ensure accurate entity recognition, which is essential for navigating the complexities of financial data and making sound investment choices 4.

Related Episodes