Introducing Smart Summaries: Generative AI for Earnings Analysis

Generative AI is changing the way people generate new ideas, collect information, and go about their day-to-day lives. It also has the potential to completely revolutionize the way we work. The future is offloading mundane, repetitive, and predictable tasks onto technology and prioritizing our time on high-value strategic and creative exercises—and that future is here. 

Today, we’re excited to announce the first of many launches powered by this technology: Smart Summaries for earnings.

AI-Generated Summarizations of Key Events

Many of our customers spend endless hours reading and summarizing earnings calls from across their portfolios, peer sets, clients, and competitors. Whether it’s scouring the Q&A of peers to prep for an earnings call, quickly understanding market-moving events for a competitive briefing, or just keeping a pulse on non-core companies that impact investment portfolios, earnings call transcripts are a key component to so many of our customers’ day-to-day tasks. 

As of today, every earnings call will have an AI-generated summary that covers the most important topics discussed in the call. Any new calls published will automatically have summaries generated within a few hours of the call itself. Summaries access the key information from a call in concise, bulleted format—without the fluff—and cite their sources. 

Smart Summaries feature:

• Bulleted summarizations that pull from the entire document and are ranked in order of importance; these will only get even smarter as people interact with them 
• Clickable citation links that bring you to the exact places within the document from which the summary was sourced
• User feedback buttons that allow you to contribute to the continuous improvement of our models

How Smart Summaries Work

Using generative AI for purposes beyond the realm of writing sonnets and having entertaining conversations is largely uncharted territory. When it comes to the use of language models trained on large swaths of the internet for decisions that affect not only individuals or businesses, but the market at large, the common (and truthfully, justified) concerns generally fall into two categories:

• Can I trust this technology to give me the correct answer? 
• Can I trust this technology with my sensitive data?

It is true that these models have a propensity for hallucination. And without a clear way to verify where data comes from, that hallucination can go dangerously undetected. Not to mention there have been several high-profile instances of data breaches of private information with consumer-grade generative AI tools. 

There are a few ways our approach to generative AI is different at AlphaSense. With our end goal being to significantly reduce the manual lift in high-value workflows, we’ve focused on building generative AI that is reliable, secure, and instantly useful. 

Artificial intelligence has always been at the core of our product offering, and as such, Smart Summaries benefits from our 10+ years of investment in AI that has been developed specifically for market intelligence. Summaries are generated using our proprietary natural language processing (NLP) technology, including semantic search (our customers might recognize this as Smart Synonyms™) and sentiment analysis, both of which are at the crux of identifying and organizing the content that should be summarized. Verticalized large language models (LLMs)—fine-tuned on our premium financial content—then consolidate those summarizations into plain language that allows for instant comprehension, without the fluff. Finally, our KPI extraction technology helps validate the accuracy of those summaries.

What does this mean in practice? 

• Higher accuracy – Leveraging AI that we’ve developed over 10+ years and trained specifically on financial data means improved accuracy and relevance and reduced risk of hallucination.
• Verifiable information. Our citations link to the snippets of language within documents that drive the summarized insights—allowing you to validate the summaries and gain trust in our model. 
• Trustworthy content – Language models rely entirely on curated content sets for the generation of summaries. AlphaSense’s model parses exclusive and coveted content sources in order to generate its high-accuracy summaries.
• Private data – Smart Summaries has been built with the most security and privacy conscious companies in mind. Even when other tools have been banned, our unique approach allows all companies to leverage the benefits of generative AI, knowing their data is safe.

Want more info about how generative AI works at AlphaSense? Check out The New Stack’s recent interview with our VP of Product Chris Ackerson. 

Expediting More Key Workflows

Looking ahead, we have an exciting roadmap when it comes to generative AI in AlphaSense. 

Up next, we will be releasing a summary experience that will live on our Company Pages, which are a critical component of competitive research, market landscaping, and portfolio management. These modules will summarize earnings performance, analyst perspectives from across our equity research collection, and key takeaways from expert interviews, all about a given company. 

After that, we will expand beyond the company-specific with portfolio and sector summaries—allowing you to take a more personalized, holistic view of the companies and industries most relevant to your research. Whether it’s a list of acquisition targets, a competitive set, or a vertical you need to watch closely, we’ll have summarized insights covering that full set of companies from across our four market perspectives to ensure you eliminate research blindspots in the portfolios or industries you cover.

All of these features will expedite the process of identifying key information and trends, providing more time savings than ever. Above all, though, we will continue to take a measured approach to ensure every piece of generative AI is ready to be used the moment it is released to you. 

Generative AI Purpose-Built for Business

For the time being, AI isn’t (and let’s face it, shouldn’t be) making strategic business decisions on its own. What AI does have the potential to do right now, though, is revolutionize what we spend our time doing in our day-to-day work. Many of the customers in our early access programs cited time savings of anywhere from 2 to 14 hours a month just from using Smart Summaries alone. And these time savings allow for more time spent on high-value tasks that only we as humans can do, like making those big-deal business decisions. 

Ready to take Smart Summaries for a spin? Start your free trial of AlphaSense today.