Hortus is dedicated to streamlining model assessments and focusing on public needs. We distill, generate, and apply AI assessment metrics that are grounded in public feedback. Here’s a sneak peek at some of our current offerings:
Reward Reports
Our documentation interface. Reward Reports are designed to track changes in model specifications, design, and reward, making it easier for you to see how different iterations perform across customized metrics.
Trajectory Trackers
Our dynamic assessment platform. Trajectory Trackers cover all of the assessment metrics recommended by San Jose’s GovAI coalition and more. The interface allows you to compare how different models would perform in ideal conditions, making procurement easier. Work with our team to develop tailored predictive metrics to see how deployed models will evolve in the ways that matter to you.
Talk to the City
Our metric design tool. Developed in collaboration with the AI Objectives Institute, Talk to the City enables feedback integration during the training process. We transform user text into model decision strategies, making sure that your AI solutions are always responsive to user needs.
More coming soon.
Hortus is continually working on ways to make the end-to-end model integration process safer, clearer, and user-oriented. Keep an eye out for more products.
Want to know more?
Learn more about the Hortus team and understand our background and mission.