Share this article
07. 05. 2024

4 min read

Insights from Semantic Visions’ Product Manager: Alyssa Parkhurst

With the BattleFin conference to be held in NYC on May 21-22 fast approaching, we took the opportunity to speak with Alyssa Parkhurst, Product Owner at Semantic Visions on the utilization of artificial Intelligence and large language models to generate alternative enriched data sets that deliver impactful and valuable insights to our clients in the investment space.

During our one-2-one discussion, Alyssa explained how Semantic Visions employs these technologies to better process and enrich OSINT data. Here’s an overview of her insights on how these innovative technologies are leveraged to deliver substantial benefits to our clients. 

Alyssa, could you explain how Semantic Visions integrates LLMs and AI into its delivery model?

Alyssa: Absolutely. At Semantic Visions, we use these technologies to significantly enhance the precision and utility of our data. We begin by collecting global news articles and other forms of media and utilizing AI-driven Named Entity Recognition to pinpoint key companies, commodities, industries,and markets, collectively defined as entities. Articles are further processed through advanced Natural Language Processing to extract hundreds of categorized events. 

These entities are correlated against identified events to create time-stamped real-world scenarios which define the entity name, industry, relevant correlated event type, relevance of the event to the entity, and the intensity of the event as it relates to the entity’s global news coverage at that time. 

The core of our technology involves using Large Language Models and Semantic Visions’ proprietary technology (LEELA) which uses algorithms that link entities to events to accurately determine contextual relevance, recognizing the role an entity plays in relation to the event and creating a summary of the instance. This refined and enriched data is then made actionable through products tailored to meet our clients’ strategic needs, such as near real-time data access, screening data feed or dashboards and detailed reports.

That sounds comprehensive. Can you describe some specific use cases that benefit your clients?

Alyssa: Of course. Our Investment offering, for example, provides enriched data feeds updated hourly to aid in dynamic trading and portfolio management. In Risk management, we leverage over a decade of expertise to offer insights into portfolio risks, supplemented by KYC/AML services for credit analysis. For the Commodity market, we track over 100 commodities, delivering insights related to market trends and potential opportunities. Lastly, our enriched data for ESG insights help clients navigate the complexities of environmental, social, and governance issues, identifying potential risks and opportunities within these frameworks.

What specific products does Semantic Visions offer to handle these needs in the alternative data landscape?

Alyssa: Semantic Visions offers a variety of products designed to transform complex data into actionable insights. Our Client dashboards provide a clear visual representation of data sets for easier decision-making. We also offer Enriched screening data feeds that allow clients to analyze trends and plan strategically. Near real-time monitoring access ensures that decisions can be made based on the latest data. ESG Score dashboards facilitate performance evaluation and industry comparison, and our Client unique reporting is tailored to meet specific client requirements, including unique data points, event types, and coverage periods.

That’s fascinating. Does Semantic Visions provide any free datasets that clients can use to backtest our data for potential alpha?

Alyssa: Yes, we provide a free backtesting dataset. It includes data from over 8,700 tickerized companies with five years of historical data across more than 400 event types. Our clients use this dataset to perform their own analysis. For example, they’ve been able to achieve a gain of 41.55%, outperforming the Nasdaq 100 by 15.56% using short selling strategies.

Thank you for this insightful interview, Alyssa. 

Alyssa: Thank you for having me! I’m really excited about the upcoming event and can’t wait to meet everyone there. It will be a great opportunity to discuss how Semantic Visions can help address the specific needs of our clients and partners. Please feel free to visit our booth at BattleFin on 21-22 May in New York. I look forward to helping you explore our solutions and discover more about what Semantic Visions has to offer.

Read next

Supplier Risk Management Best Practices: 6 Steps to Safeguard Your Business from Supply Chain Disruptions
05. 02. 2025 - 18 min read

Supplier Risk Management Best Practices: 6 Steps to Safeguard Your Business from Supply Chain Disruptions

Introduction: How Semantic Visions Transforms Supplier Risk Management Managing supplier risks is not just about avoiding disruptions—it’s about gaining a competitive advantage. In my experience, the companies that thrive are those that leverage real-time intelligence...

Read more
USA Today – Revolutionizing Supply Chains: How Semantic Visions is Transforming Risk Management with AI
04. 02. 2025 - 4 min read

USA Today – Revolutionizing Supply Chains: How Semantic Visions is Transforming Risk Management with AI

This article was originally published in USA Today, where Jan Balatka, Managing Director of Semantic Visions, discusses how Semantic Visions is transforming AI supply chain management with advanced risk detection and data-driven insights. As global...

Read more
No Small Beer: Big Data in Tackling Global Supply Chains
30. 01. 2025 - 7 min read

No Small Beer: Big Data in Tackling Global Supply Chains

You may have heard of the Beergame and the lessons it tries to teach. Developed in the 1960s at the MIT Sloan School of Management, for decades it has provided “a laboratory for exploring how...

Read more

Ready to see it in action?

Let's meet