Skip links

Built-in Foundational Models for the Financial Sector

A Look at Sujet AI Products


In the rapidly evolving world of finance, artificial intelligence (AI) is becoming a game-changer. One of the most promising applications of AI is the use of foundational models, which are deep learning architectures designed to understand and generate human-like text or multimodal content. These models, such as instruct-based Language Language Models (LLMs) and vision language models (VLMs), can be fine-tuned for specific financial tasks, offering numerous benefits such as privacy, personalization, and high performance. In this post, we’ll explore the importance of built-in foundational models for the financial sector and introduce Sujet AI’s innovative products.

Understanding Foundational Models and Their Categories

Foundational models are AI models that are trained on broad data at scale and can be adapted (e.g., fine-tuned) to a wide range of downstream tasks. The categories of foundational models include instruct-based LLMs, VLMs, Retrieval-Augmented Generation (RAG), text embedding, and agents.

In the finance sector, these models can be used in various ways. For instance, instruct-based LLMs can be fine-tuned to understand and generate financial reports or even predict financial price shares. VLMs, on the other hand, can be used to analyze multimodal data, such as images and text from news articles, to predict stock market trends. RAG can be used to provide more accurate and detailed financial information, while text embedding can help in clustering similar financial documents or reports. Agents can be used to automate financial tasks, such as trading or portfolio management.

The Benefits of Built-in Models

Built-in models offer numerous benefits for the financial sector. Firstly, they provide a higher level of privacy and security as they can be fine-tuned on a company’s private data without the need to share it with third-party providers. Secondly, they allow for greater personalization as they can be fine-tuned to a company’s specific needs and requirements. Thirdly, they offer high performance as they are specifically designed and fine-tuned for financial tasks.

Sujet AI is at the forefront of developing and open-sourcing built-in foundational models for the financial sector. We have open-sourced two datasets on Hugging Face: sujet-ai/Sujet-Finance-Instruct-177K and sujet-ai/Sujet-Finance-Vision-10K, which are among the largest datasets available.

We have also open-sourced an instruction-based model, sujet-ai/Sujet-Finance-8B-v0.1, specifically designed for financial tasks. This model outperforms the Llama 3 8B model, developed by Meta AI, with a high margin. Our model is available on Hugging Face, a popular platform for training and deploying AI models.

Applications of Sujet AI's Models in the Financial Sector

Sujet AI’s models can be used in a wide range of financial applications, from investment and raising capital to venture capitals and equity trading. For instance, our models can be fine-tuned to assist financial consultants and analysts in making informed investment decisions or to help small businesses and charities raise capital through crowdfunding or savings accounts.

Our models can also be used by venture capitals to analyze and choose the best companies for their portfolio or by equity traders to predict market trends and make profitable trades. Moreover, our models can be used by hedge funds, crypto, midcap, and mutual funds to automate and enhance their financial tasks.


In the ever-evolving landscape of the financial sector, built-in foundational models are set to play a crucial role. They offer numerous benefits, such as privacy, personalization, and high performance, and can be used in a wide range of financial applications. Sujet AI is proud to be at the forefront of this revolution, developing and open-sourcing innovative built-in models for the financial sector. We invite you to explore and leverage our models to enhance and automate your financial tasks.

Leave a comment

🍪 This website uses cookies to improve your web experience.
Stay In Touch

Be the first to know about our news and events