Our journey into the world of investment and financial applications began in 2014.
We developed ultra-low latency applications, built derivative pricing models, worked on exchange connectivity, and helped hedge funds and banks automate regulatory reporting and accurately reflect these in their pricing.
Since the "ChatGPT moment" in late 2022, we've been exploring how AI and its modern implementations can be applied to investment strategies. After 18 months of intensive research and dozens of completed projects, we've decided to focus entirely on AI applications.
Here's why.
AI's Current Limitations
The present level of AI development, particularly in large language models (LLMs), has significant constraints. While reasoning is often touted as a key metric, we believe it's premature to rely on this capability.
In our view, decision-making through LLMs remains more utopian than practical. These models struggle with numerical operations, often making mistakes in basic arithmetic and comparative tasks, especially with extensive context.
You might wonder: Aren't calculation and decision-making core skills in investment?
AI's Strengths in Finance
Indeed, it's hard to imagine a successful investor who can't calculate or make decisions. These skills are still crucial for human investors.
However, LLMs excel at structuring unstructured information. In the investment world, information is the most valuable resource. More information than your competitors means more accurate decisions.
News, press releases, opinions, rumors, social media discussions, product descriptions, reviews, and scientific publications are all sources of information. But they're poorly structured.
LLMs are tools that can extract crucial data from disparate texts. Models previously used to analyze prices, revenues, EPS, and other numerical indicators now have access to an unprecedented amount of new data.
AI's Additional Capabilities
While the previous argument alone justifies the importance of applying LLMs in investment applications, we'd like to highlight another significant talent of these models.
It's impossible to interpret any fact in isolation. What matters in investment decision-making is how a fact will influence various factors. The number of interconnections between facts and events is in the millions - far beyond the capacity of even the most talented analyst to comprehend and update in real-time based on global events.
LLMs make this possible.
We see enormous potential here for creating a new generation of analytical tools.
Our Future Plans
While the previous argument alone justifies the importance of applying LLMs in investment applications, we'd like to highlight another significant talent of these models.
As mentioned, we've decided to focus entirely on creating customized AI-based solutions for our clients in the investment application sphere.
Specifically, we're developing:
We developed ultra-low latency applications, built derivative pricing models, worked on exchange connectivity, and helped hedge funds and banks automate regulatory reporting and accurately reflect these in their pricing.
Since the "ChatGPT moment" in late 2022, we've been exploring how AI and its modern implementations can be applied to investment strategies. After 18 months of intensive research and dozens of completed projects, we've decided to focus entirely on AI applications.
Here's why.
AI's Current Limitations
The present level of AI development, particularly in large language models (LLMs), has significant constraints. While reasoning is often touted as a key metric, we believe it's premature to rely on this capability.
In our view, decision-making through LLMs remains more utopian than practical. These models struggle with numerical operations, often making mistakes in basic arithmetic and comparative tasks, especially with extensive context.
You might wonder: Aren't calculation and decision-making core skills in investment?
AI's Strengths in Finance
Indeed, it's hard to imagine a successful investor who can't calculate or make decisions. These skills are still crucial for human investors.
However, LLMs excel at structuring unstructured information. In the investment world, information is the most valuable resource. More information than your competitors means more accurate decisions.
News, press releases, opinions, rumors, social media discussions, product descriptions, reviews, and scientific publications are all sources of information. But they're poorly structured.
LLMs are tools that can extract crucial data from disparate texts. Models previously used to analyze prices, revenues, EPS, and other numerical indicators now have access to an unprecedented amount of new data.
AI's Additional Capabilities
While the previous argument alone justifies the importance of applying LLMs in investment applications, we'd like to highlight another significant talent of these models.
It's impossible to interpret any fact in isolation. What matters in investment decision-making is how a fact will influence various factors. The number of interconnections between facts and events is in the millions - far beyond the capacity of even the most talented analyst to comprehend and update in real-time based on global events.
LLMs make this possible.
We see enormous potential here for creating a new generation of analytical tools.
Our Future Plans
While the previous argument alone justifies the importance of applying LLMs in investment applications, we'd like to highlight another significant talent of these models.
As mentioned, we've decided to focus entirely on creating customized AI-based solutions for our clients in the investment application sphere.
Specifically, we're developing:
- Analytical tools for news processing, tailored for portfolio managers in hedge funds, family offices, and professional and retail investors
- Automated company report analysis for both value and growth investors
- Personalized newsletters for day traders and long-term investors alike
MORE ARTICLES