Intrinio recently launched Thea, an artificial intelligence engine that can comprehend large volumes of text rapidly, with the accuracy of a human. When it comes to data like SEC filings, Thea is a gamechanger for investors searching for alpha.
Intrinio’s data engineers built Thea by leveraging extremely recent advances in Natural Language Processing. Thea runs off an artificial neural network that processes data at the level of a college graduate. A neural net consists of layers of nodes that each receive data and send it to the nodes around it.
A weight and vectorization system determines when the data is output by the system, and essentially acts as a filtering device to eliminate noise. So, when a user asks for environmental challenges, they will specifically get environmental challenges, not just all challenges. When the answer is returned, it consists of a smaller portion of text specific to the question, as well text for context, and a link to the document it originated from.
These answers are delivered via our Answers API, which allows users to build a UI (such as a search box) to retrieve answers or query the API directly and leverage the answers in deep learning models and cluster analyses.
Thea currently has access to SEC filings going almost 30 years back, ranging from 10-Ks and 10-Qs to special disclosures.
However, Thea is dataset agnostic, meaning that it can process and return points from any dataset that is NLP-friendly (essentially, most textual documents). We can integrate any natural language dataset, such as internal reports, legal documents, or any other documents a user wants to use for research. The applications are virtually endless, and in the world of big data, an engine like Thea is incredibly powerful.
While it depends on the dataset, Thea is able to answer any question that a human would be able to answer from scanning the text. We tried it out ourselves, asking Thea what sustainability challenges S&P 500 companies are facing and what business strategies S&P 500 companies leverage, for example.
Thea is highly efficient, doing all the research and mining work for you. Hedge funds, brokerages, and fintechs can use Thea and the Answers API for investment research in several ways:
Even law firms, journalists, and academics can use the Answers API. To dive deeper into potential use cases, check out our blog about the Answers API.
Interested in learning more about Thea and the Answers API? Request a consultation.