
I've built a pipeline that takes human-written stories and generates metadata that allows a small model to learn how to tell novel stories in the way a human does. The metadata generated includes likely instructions from a human that would lead to generation of a given progression in the story, the likely internal thoughts that the writer would have as they wrote the progression of the story, and more. The metadata is then formatted in input/output pairs, and the model is trained on those pairs using SFT. The resulting model demonstrates a much more natural writing style than most AI models, and is highly steerable
1 Dec 2024
.png&w=828&q=75)
Social services often face limitations in manpower and resources, making it challenging to consistently check in with clients and track progress on action plans. While these plans offer comprehensive goals and resources, they lack manageable, step-by-step tasks to guide clients effectively toward their goals. Our solution addresses this gap by transforming these extensive goals and resource documents into actionable tasks, broken down into manageable steps distributed over a specific period. Additionally, a proactive AI voice agent regularly checks in with clients, tracking their progress and updating a dashboard accessible to social workers. This enables social workers to stay informed, monitor client progress in real-time, and make data-driven decisions to offer better support.
11 Nov 2024