Last week was an incredible week for OpenBB.
On Monday, we launched Product Hunt, where we won Product of the Day and were featured in an article by TechCrunch: Fintech OpenBB aims to be more than an ‘open source Bloomberg Terminal’”.
PyQuantNews posted a thread on X about OpenBB that generated 1.3 million impressions: PyQuant News 🐍 on Twitter / X.
We hosted a webinar with hundreds of OpenBB community members, which you can find here: Webinar Recording - The Next Chapter: AI Meets the OpenBB Terminal
And to conclude the week, we crossed 60,000 registered users.
In case you missed last week’s activity — this post explains our OpenBB vision, our journey so far, and where we are headed.
Context
When I started working on Gamestonk Terminal, I had no idea of the impact that this side project would have.
While I was tinkering in the right direction — a combination of integrating multiple datasets within one interface and automating my investment research process — I was missing an integral piece of the puzzle.
I needed AI to take my workflow to the next level.
When ChatGPT launched, we understood the impact of this technology and how it would change everything. However, we never believed that a chat would be the right interface to do research. If you're curious to know why, I wrote more about it here: Why chat-only AI Financial Assistants are not the future.
So whilst we watched as hundreds of startups built an “SEC filings chatbot”, “PE data chatbot”, and “RAG on your data”, we decided to go in a different direction…
We compared the technology in use across many types of financial users, including:
Desktop providers as the interface to perform investment research like Bloomberg, Refinitiv, and FactSet
Cloud vendors marketplaces that offer access to a vast range of financial data like Snowflake, Databricks, and AWS
Analytics tools that enable users to bring their own data and create custom visualizations like Streamlit, Tableau, and PowerBI
LLMs that can make analysts/quants workflow more efficient (summarization, extracting patterns from data, Q&A) like OpenAI, Anthropic, and Gemini
And we couldn’t find a solution that met the conditions we wanted - an AI-powered research and analytics workspace. For almost two years, we put our heads down and built a product that has the power to be transformative for financial research. A product that sits at the intersection of these solutions.
We also assumed that the models would keep getting better, to the point where agentic workflows would become more and more reliable. Agents perceive their environment, act, observe the results, and adjust. This ongoing process and feedback loop helps them adapt to real-world changes.
This is not just a financial terminal with a chatbot on the side. This is a full-blown Compound AI System (the term was coined by Berkeley University earlier this year - The Shift from Models to Compound AI Systems).
Data
It all starts with data.
We wanted to offer access to data from many data vendors and data marts. However, if we stopped there, it would be an incomplete offering. So we engineered the platform for users to be able to integrate any data from any source. Whether the data is structured (CSV, json, xlsx) or unstructured (PDF, images, or e-mail), a website or an API endpoint, or a proprietary backend, users can pipeline all the data they need to perform research and analytics.
We have also open-sourced the backend data layer so users can have full control over the data they want to bring into the product. And, in addition, we also have native database integrations like MySQL, SQLite, Snowflake, and others.
Visualization
After hundreds of conversations with analysts, quants, investors, engineers, and students, we understood the gaps in their tooling.
They wanted to have all their data available within the same product and be efficient in their analysis. More importantly, they wanted to be able to share that data with others, both through sharing a dashboard and a report.
The one-size-fits-all solution didn’t meet users’s needs.
Ultimately, they wanted flexibility.
So we built our interface to provide unparalleled flexibility and ensure that users could control their entire workflows.
AI Insights
And finally, the missing ingredient that ties everything together.
We have built the OpenBB Copilot for over one year. It has dozens of different workflows that ultimately result in an AI Copilot that has better accuracy and less hallucinations compared to industry benchmarks - more on this in the future.
The OpenBB Copilot has agentic capabilities, and over time it will be able to do more and more complex workflows based on the type of data available and the better foundational LLMs available.
We are building a future where you can ask the OpenBB Copilot, “What are the biggest risks of investing in X?” and it will be able to create a dashboard with dozens of widgets and relevant data for such assets. It will provide a full investment thesis on that asset based on the data, being able to spot patterns and trends that would take hours or days to an analyst - but in a fraction of the time.
See it in action with this earnings workflow: How to analyze Earnings with Copilot on OpenBB Terminal
Bring Your Own Copilot
Given that many of our users are from firms where security and privacy are a must, we have built a framework that enables firms to bring their own copilots. Whether it’s a fine-tuned model on their own internal data or if it does RAG within their database - everything is possible.
We have open-sourced that repository so that AI engineers and developers can integrate their own LLMs into our AI-powered research and analytics workspace. See it here: GitHub - OpenBB-finance/copilot-for-terminal-pro: Copilot template to bring your own copilots into the OpenBB Terminal Pro.
Products for Individuals and Professionals
Whilst we’ve built the OpenBB Terminal Pro for almost two years, we have kept this product focused on enterprise use cases.
There were two main reasons for this:
We wanted to build a product that was enterprise-ready
We wanted to focus on specific use cases and workflows
Now that we’ve achieved that, we are ready to welcome the rest of our community. Over the past three years, we have amassed a supporting community of hundreds of thousands of investors, analysts, and quants across the globe.
Most people in the financial world have heard about OpenBB, yet the OpenBB Terminal Pro hasn’t been accessible to them.
Until today.
We are, for the first time, creating a free tier for our AI-powered research and analytics workspace. And we are calling it: “OpenBB Terminal”.
We’re just getting started, and I can’t wait to hear your feedback on the new OpenBB Terminal.
Sign up for our newsletter so you don’t miss any future announcements, blog posts, and webinars.