AI Readiness in Finance
Trusted and efficient financial AI requires robust benchmarks, models, agents, and deployment. The Fin AI collaborates with industry and academia to continuously enhance accuracy, reliability, and real-world impact.
What We Do
Dataset & Benchmark
Reliable financial AI relies on comprehensive datasets for pre-training, fine-tuning, and evaluation. Holistic benchmarks provide transparent and consistent performance measurement. These tools drive innovation and ensure trustworthy AI development.
Model
Open FinLLMs introduces FinLLaMA, FinLLaMA-instruct, and FinLLaVA, financial LLMs pre-trained on a 52B-token corpus and fine-tuned with 573K text and 1.43M multimodal instructions. Released with checkpoints and training scripts, they advance financial AI accessibility and innovation.
Agent
Developing a multi-agent system with human-like cognition and dynamic risk profiling to enhance multi-turn financial decision-making. Using multimodal financial data, it bridges intelligent system design with behavioral finance for diverse financial applications.
Trustworthy Deploy
Creating a federated learning-based application framework for reliable and trustworthy financial LLM applications, emphasizing privacy and security in financial AI deployments.
COMMUNITY
Bridging AI and Finance Through People
We are a strong community of researchers, engineers, and financial experts is the driving force behind bridging AI and finance. It’s their collaboration, expertise, and dedication that advance trustworthy and impactful financial AI systems.
LATEST UPDATES