In this episode of Leading with Data, we feature Didier Rodrigues Lopes, the Founder and CEO of OpenBB, a trailblazer in the field of AI-powered research and analytics. With over 34K GitHub stars in under three years, Didier shares his journey of transforming a personal project into a thriving open-source business that empowers financial research and data analysis.
You can listen to this episode of Leading with Data on popular platforms like Spotify, Google Podcasts, and Apple. Pick your favorite to enjoy the insightful content!
Let’s look into the details of our conversation with Didier Rodrigues Lopes!
Yeah, sure. My journey into finance began when my former university math teacher in Portugal, who was working on his PhD thesis on modeling and forecasting financial time series, reached out to me. He challenged me to write the code for his thesis, which introduced me to the financial space. After working with him for about eight months, I started doing my own investment research. I quickly realized how cumbersome the process was, especially when it came to data retrieval. As an engineer, I couldn’t understand why I couldn’t automate this process since the data format was the same regardless of the asset. This frustration led me to create my own financial terminal during a week when my flight got canceled due to COVID. I looked for open-source projects on GitHub to build upon, but finding none, I started my own, which eventually became OpenBB.
The initial response was overwhelming. I expected around 20 GitHub stars but hit 4,000 within 24 hours. The community formed rapidly, and I needed a space to discuss the project’s direction. We created a Discord server that became our community’s central hub.
People quickly started adding features and improving the code. I learned a lot from both engineers and financial professionals about their data needs. For instance, I hadn’t initially included options data, but community members soon contributed that feature. Someone else added a full Forex menu.
It was incredible to see how people’s perspectives and data interests could shape the project’s development. Each contribution helped me understand different workflows and user needs.
The transition happened when we were approached by OSS Capital. We discussed the future of the product and why it had to be open source to work. The capital we raised was to accelerate our vision and to figure out how to make money during that time. Our vision was to create an open ecosystem for research, where users could bring their own data and build any type of features on top of it. Unlike Bloomberg or FactSet, which are high-quality but limited in customizability, we wanted to offer a fully customizable product that could only be achieved through open source. This meant users could fully customize the code and the product to their needs.
The rebranding was necessary because the original name, Gamestonk Terminal, was too closely associated with GameStop and was seen as meme-like. We wanted a name that would be taken seriously. We agreed on “Open” because it represented what we stood for, and “BB” came from a personal connection – both my co-founder and I were investors in BlackBerry and were losing money on it. So, we named the company OpenBB, which also allowed us to play with the butterfly logo design.
Currently, OpenBB offers an open-source terminal focused on data aggregation and standardization, which is free for individuals. We also have a web-based version for financial firms that allows for collaboration and research at a low cost. Our roadmap includes making it even easier to bring any type of data into the product, expanding the capabilities of our custom backend, and enhancing our copilot feature. We want to open up the architecture even more so that users can connect their own large language models to OpenBB, making it a router towards the best model for the task at hand.
It’s a big problem in the industry, and we don’t claim to have zero hallucinations because that’s not realistic. Instead, we focus on systems that allow users to validate the accuracy of the AI’s responses. Our citation system tells users which widgets were used to answer a question, allowing them to validate the original source. We also have internal systems to minimize hallucinations, but it’s more about providing tools for users to quickly validate the data.
It varies day by day. Mornings are usually filled with meetings due to time zone overlaps with our European team. Afternoons are for focused work, which could be anything from implementing a feature to visiting a company for a product demo. At night, I try to code because I enjoy it and can often find small improvements to make to the product. However, as the product has grown and uses technologies like React, which I’m less familiar with, it takes more time to get features ready.
On weekends, I often stay home and work, though I sometimes play soccer on Saturdays. For holidays, I enjoy visiting new countries or cities, learning about their culture, and understanding their history. Recently, I visited Chicago and enjoyed tours to understand the city’s architecture and history.
Didier Rodrigues Lopes’s journey from a math student to a tech entrepreneur exemplifies how personal frustration can spark technological innovation. By prioritizing open-source principles and user customization, he has created a platform that challenges traditional financial research tools and democratizes access to sophisticated analytics.
For more engaging sessions on AI, data science, and GenAI, stay tuned with us on Leading with Data.