R&D experiment. It can change or be removed at any time and is not reviewed or approved. Do not rely on it for decisions. Read the full warning.
R&D / Experiment
What can AI be evaluated on in banking?
Explore banking capabilities, suggested evaluation tasks and data sources. The catalogue describes evaluation designs; it does not report test results.
New here? How the library works, a five-step guide.
Banking capabilities
Dodone's own topic groups, with links to BIAN v14.0 service domains that are checked against the source.
6 topics
Suggested evaluations
Specifications for testing an agent, a model or a retrieval step on one banking task.
6 published specifications
Datasets and sources
Public, synthetic and restricted sources that could support an evaluation, with their limits.
8 published source cards
The BIAN v14.0 landscape, for scale
BIAN's own totals for its whole v14.0 landscape. They are not the size of this library.
- 322
- 38
- 586
- 4,905
- 362
- 242
Comments
Tell us what is useful, what is wrong and what is missing in this experiment.
No published comments yet.
Sign in to comment.