otai-hf-local
Freeby Softoftcompany
Hugging Face plugin for Open Ticket AI enabling local, privacy-preserving ticket classification using machine learning.
otai-hf-local is a specialized plugin for the Open Ticket AI (OTAI) framework. It enables the integration of Hugging Face Transformers directly into the ticket workflow of open-source ITSM systems. By running machine learning models locally, support tickets can be automatically categorized, prioritized, or analyzed for sentiment while maintaining full data sovereignty. The plugin is particularly relevant for organizations with high compliance requirements that want to leverage AI automation without transferring sensitive ticket data to the cloud. It works seamlessly with the OTAI connectors for OTOBO, Znuny, and Zammad.
Features
Local Inference
Runs NLP models entirely on your own infrastructure without sending data to external APIs.
Hugging Face Integration
Supports any Hugging Face text classification model for categorization, priority detection, and sentiment analysis.
GPU Acceleration
Optional CUDA support for faster processing in high-volume ticket environments.
Confidence Scores
Provides probability scores for predictions to support threshold-based automation.