Fundamental Research Engineering (FRE):
Redefining Investment Research
with Generative AI
Redefining
Investment Research
with Generative AI
Redefining
Investment Research
with Generative AI
Forbes 30 Under 30 Asia (AI)
— Highlighted Honoree
Forbes 30 Under 30 Asia (AI)
— Highlighted Honoree
“I’ve seen firsthand how time-consuming and frustrating traditional investment research can be. There was always a gap between the data available and the insights analysts needed. Now, we’re building the solution I wish existed—faster, smarter, and built for real-world investing.”
“I’ve seen firsthand how time-consuming and frustrating traditional investment research can be. There was always a gap between the data available and the insights analysts needed. Now, we’re building the solution I wish existed—faster, smarter, and built for real-world investing.”
Jin Kim
Jin Kim
Head of Fundamental Research Engineering
Head of Fundamental Research Engineering
Bridging AI and Finance
Bridging AI and Finance
At LinqAlpha, we believe the most powerful AI solutions come from deep domain expertise. That’s why our Fundamental Research Engineering (FRE) team brings together world-class AI talent with seasoned professionals from the front lines of finance—across research, sales, and trading on both the buy and sell side.
At LinqAlpha, we believe the most powerful AI solutions come from deep domain expertise. That’s why our Fundamental Research Engineering (FRE) team brings together world-class AI talent with seasoned professionals from the front lines of finance—across research, sales, and trading on both the buy and sell side.
We don’t just build AI—we build with firsthand knowledge of what finance professionals actually need. From stock screening and rapid primers to gauging macro sentiment and more, we’ve lived the workflows we’re automating. This fusion of technical depth and domain fluency enables us to engineer AI systems that tackle real problems—not theoretical ones.
We don’t just build AI—we build with firsthand knowledge of what finance professionals actually need. From stock screening and rapid primers to gauging macro sentiment and more, we’ve lived the workflows we’re automating. This fusion of technical depth and domain fluency enables us to engineer AI systems that tackle real problems—not theoretical ones.
What is FRE?
The Problem
The Problem
We know that legacy tools often fall short—whether it’s identifying comparable peers in a new industry or synthesizing insights from management commentary, filings, and research notes. The problem is not “lack” of novel data, but the inability of legacy software to read “narratives” told by existing data. That’s why we’ve built FRE to go beyond traditional quant methods, uncovering the qualitative narratives behind the numbers and delivering insights that are both bespoke and scalable.
We know that legacy tools often fall short—whether it’s identifying comparable peers in
a new industry or synthesizing insights from management commentary, filings, and research notes. The problem is not “lack” of novel data,
but the inability of legacy software to read “narratives” told by existing data.
That’s why we’ve built FRE to go beyond traditional quant methods, uncovering
the qualitative narratives behind the numbers
and delivering insights that are both bespoke
and scalable.
The Solution
The Solution
Take stock screening: Legacy tools often fall short in identifying comparable peers in new industries, leaving investors to manually verify and customize lists. The challenge isn’t a lack of data—it’s the inability of traditional systems to interpret the rich insights embedded in management commentary, filings, channel checks, and research notes.
Take stock screening: Legacy tools often fall short in identifying comparable peers in new industries, leaving investors to manually verify and customize lists. The challenge isn’t a lack of data—it’s the inability of traditional systems to interpret the rich insights embedded in management commentary, filings, channel checks, and research notes.
In essence, FRE combines deep domain expertise with GenAI-driven, agentic workflows, allowing LinqAlpha to translate fundamental investors’ bottom-up perspectives into dynamic and scalable intelligent systems.
In essence, FRE combines deep domain expertise with GenAI-driven, agentic workflows,
allowing LinqAlpha to translate
fundamental investors’ bottom-up perspectives
into dynamic and scalable intelligent systems.
Objective
Objective
Input Data
Input Data
Methodology
Methodology
Target Output
Target Output
Time Horizon
Time Horizon
Quantitative Research Engineering
Quantitative Research Engineering
Identify and exploit immediate market inefficiencies through patterns and signals
Identify and exploit immediate market inefficiencies through patterns and signals
Large volumes of numerical, measurable data from systematic pipelines
Large volumes of numerical, measurable data from systematic pipelines
Statistical & regression-based modeling with data updates and model recalibration
Statistical & regression-based modeling with data updates and model recalibration
Actionable trading signals, factor exposures, and forecasts
Actionable trading signals, factor exposures, and forecasts
Short-term & may degrade as signals become widely known
Short-term & may degrade as signals become widely known
Fundamental Research Engineering
Fundamental Research Engineering
Curate rich, context-driven narratives to enhance human judgment
Curate rich, context-driven narratives to enhance human judgment
Variable volumes of non-numerical, textual data tailored to investor needs
Variable volumes of non-numerical, textual data tailored to investor needs
Personalized to each fund’s strategy, evolving with instructions and feedback
Personalized to each fund’s strategy, evolving with instructions and feedback
Agentic analysis of qualitative and quantitative information
Agentic analysis of qualitative and quantitative information
Both short-term catalysts and mid to long-term positioning
Both short-term catalysts and mid to long-term positioning
Objective
Objective
Input Data
Input Data
Methodology
Methodology
Target Output
Target Output
Time Horizon
Time Horizon
Quantitative Research Engineering
Quantitative Research Engineering
Identify and exploit immediate market inefficiencies through patterns and signals
Identify and exploit immediate market inefficiencies through patterns and signals
Large volumes of numerical, measurable data from systematic pipelines
Large volumes of numerical, measurable data from systematic pipelines
Statistical & regression-based modeling with data updates and model recalibration
Statistical & regression-based modeling with data updates and model recalibration
Actionable trading signals, factor exposures, and forecasts
Actionable trading signals, factor exposures, and forecasts
Short-term & may degrade as signals become widely known
Short-term & may degrade as signals become widely known
Fundamental Research Engineering
Fundamental Research Engineering
Curate rich, context-driven narratives to enhance human judgment
Curate rich, context-driven narratives to enhance human judgment
Variable volumes of non-numerical, textual data tailored to investor needs
Variable volumes of non-numerical, textual data tailored to investor needs
Personalized to each fund’s strategy, evolving with instructions and feedback
Personalized to each fund’s strategy, evolving with instructions and feedback
Agentic analysis of qualitative and quantitative information
Agentic analysis of qualitative and quantitative information
Both short-term catalysts and mid to long-term positioning
Both short-term catalysts and mid to long-term positioning
The FRE Team
The FRE Team
Jaeseon Ha
Ex-Goldman Sachs TMT Analyst
"As a TMT analyst, I spent hours breaking down earnings reports, tracking industry trends, and manually adjusting models. The research process often felt reactive rather than proactive. Now, we’re building AI-powered workflows that surface critical insights in real time, helping analysts stay ahead of market-moving events and focus on higher-value investment research."
Fundamental Research Engineer, Domain Expert
Jaeseon Ha
Ex-Goldman Sachs TMT Analyst
"As a TMT analyst, I spent hours breaking down earnings reports, tracking industry trends, and manually adjusting models. The research process often felt reactive rather than proactive. Now, we’re building AI-powered workflows that surface critical insights in real time, helping analysts stay ahead of market-moving events and focus on higher-value investment research."
Fundamental Research Engineer, Domain Expert
Jaeseon Ha
Ex-Goldman Sachs TMT Analyst
"As a TMT analyst, I spent hours breaking down earnings reports, tracking industry trends, and manually adjusting models. The research process often felt reactive rather than proactive. Now, we’re building AI-powered workflows that surface critical insights in real time, helping analysts stay ahead of market-moving events and focus on higher-value investment research."
Fundamental Research Engineer, Domain Expert
Suyeol Yun
MIT MS
"Traditional research tools struggle to make sense of unstructured data, leaving analysts to manually piece together fragmented insights. We’re developing AI agents that don’t just retrieve information but interpret and structure it, allowing investors to navigate earnings calls, filings, and reports with precision. Our goal is to make fundamental research more dynamic, personalized, and efficient."
Fundamental Research Engineer, LLM Expert
Suyeol Yun
MIT MS
"Traditional research tools struggle to make sense of unstructured data, leaving analysts to manually piece together fragmented insights. We’re developing AI agents that don’t just retrieve information but interpret and structure it, allowing investors to navigate earnings calls, filings, and reports with precision. Our goal is to make fundamental research more dynamic, personalized, and efficient."
Fundamental Research Engineer, LLM Expert
Suyeol Yun
MIT MS
"Traditional research tools struggle to make sense of unstructured data, leaving analysts to manually piece together fragmented insights. We’re developing AI agents that don’t just retrieve information but interpret and structure it, allowing investors to navigate earnings calls, filings, and reports with precision. Our goal is to make fundamental research more dynamic, personalized, and efficient."
Fundamental Research Engineer, LLM Expert
Subeen Pang
MIT CSE Ph.D
"AI should do more than automate tasks—it should enhance human decision-making. At LinqAlpha, we’re building intelligent research workflows that adapt to complex financial data, allowing analysts to focus on deeper analysis rather than manual data aggregation. By integrating agentic workflows with high-quality proprietary data, we make fundamental research faster, more scalable, and more precise."
Fundamental Research Engineer, LLM Expert
Subeen Pang
MIT CSE Ph.D
"AI should do more than automate tasks—it should enhance human decision-making. At LinqAlpha, we’re building intelligent research workflows that adapt to complex financial data, allowing analysts to focus on deeper analysis rather than manual data aggregation. By integrating agentic workflows with high-quality proprietary data, we make fundamental research faster, more scalable, and more precise."
Fundamental Research Engineer, LLM Expert
Subeen Pang
MIT CSE Ph.D
"AI should do more than automate tasks—it should enhance human decision-making. At LinqAlpha, we’re building intelligent research workflows that adapt to complex financial data, allowing analysts to focus on deeper analysis rather than manual data aggregation. By integrating agentic workflows with high-quality proprietary data, we make fundamental research faster, more scalable, and more precise."
Fundamental Research Engineer, LLM Expert