In Finrobot, there are three folders: agents, data_source, and functional, along with utils and toolkit Python files. agents library include from finrobot.data_source import * from finrobot.functional import * from textwrap import dedent library = [ { "name": "Software_Developer", "profile": "As a Software Developer for this position, you must be able to work collaboratively in a … Continue reading Reference FinRobot Building Auto Agents

CrewAI and AutoGen (pyautogen) are both frameworks for building AI agents, but they have distinct approaches and features: Ease of Use CrewAI: More accessible and easier to set up, built on LangChain2. AutoGen: May require more effort to set up initially, but offers more flexibility for specialized tasks2. Functionality CrewAI: Provides more control over the … Continue reading Compare CrewAI and Auto Agent

These two papers showcase how it's feasible to build an AI platform replacing a whole team of financial analysts: FinRobot: An Open-Source AI Agent Platform for Financial Applications using Large Language Models (github repo) and A First Look at Financial Data Analysis Using ChatGPT-4o. Even these papers only provide prompts for reference purpose, the true … Continue reading AI Platforms Replacing Financial Analysts

Given the current trend, hedge funds are likely working diligently to leverage artificial intelligence in their stock selection processes. It seems prudent to apply similar approaches to democratized index strategies for exchange-traded funds (ETFs). A pioneer in this field, Qraft Technologies, is dedicated to democratizing quantitative investing by utilizing AI models for portfolio construction and … Continue reading ETFs Constructed by AI Technology, Pioneered by Qraft

MongoDB can store data as documents (usually in JSON format, specifically BSON – Binary JSON).Each document can have varying fields and structures, meaning records don't need to follow a strict schema. Good for unstructured or semi-structured data (e.g., user profiles, logs, or product catalogs). While Relational SQL Databases (RDBMS) store data as table-based, Requires a … Continue reading MongoDB/NoSQL versus Relational SQL Databases (RDBMS)

Due to the rapid growth of artificial intelligence, there has been a significant increase in demand for Graphics Processing Units (GPUs). Each GPU operates with its own hardware-level language known as CUDA. While an in-depth understanding of CUDA is not strictly necessary, a foundational grasp of its principles is essential for effectively applying AI methodologies. … Continue reading Why GPU and Necessity to Learn CUDA?

After conducting tests on the U.S. market using both large-cap and small-cap proxies, I found that the passive ownership factor holds true. Specifically, companies with higher passive ownership—such as ETFs and mutual fund holders—tend to exhibit stronger stock performance. This effect is both persistent and more pronounced within the small-cap universe. Building on these findings, … Continue reading Passive Factor Analysis on Taiwan TAIEX 16 Years History

If I have various versions of a file based on one major framework, I can create multiple Git branches to track and maintain them without creating redundant messy copies all over the place. For example, on my passive_factor repo, the main branch is the framework, where I tested out the passive factor does exist in … Continue reading Comparing Git and Microsoft Versioning

The biggest take away from the latest talk by Jensen Huang in AI summits is concepts of "super employee" who can leverage a suite of AI agents to do the virtual work and "physical AI", i.e. humanoid who can literally replace a human. as a premise, he mentioned "blackwell", Nvidia's Blackwell refers to a new generation of … Continue reading Super-employee and Physical AI, Humanoid

To learn a great deal from Y Combinator and Peter Koomen regarding enterprise sales, particularly for startups selling software to enterprises, it's crucial to understand the unique dynamics of these transactions. Enterprise sales can be multifaceted, often involving longer sales cycles, multiple stakeholders, and the need for tailored solutions. By drawing insights from experts, one … Continue reading Enterprise Sales Process