Streamlining Information Search Using Gen AI

System Research in Japan used generative AI tools IBM Watson Discovery and IBM watsonx.ai to improve information search across internal documents. By enabling natural language queries and tuning the AI-powered search interface, the company reduced the time needed to find required information by about 50%, uncovering specific details more efficiently and enhancing overall retrieval accuracy.

Deploying Models Coverted to ONNX Format

The article explains how developers can deploy machine learning models that have been converted to the ONNX (Open Neural Network Exchange) format, which is an open-source standard for representing models from frameworks like PyTorch and TensorFlow. It highlights the benefits of ONNX such as improved cross-platform compatibility, performance optimizations, and interoperability and outlines basic steps for converting models and deploying them using ONNX runtime and deployment tools.

Supported Foundation Models In Watsonx.AI

IBM watsonx.ai offers various foundation models for generative AI, including IBM‑curated, custom, and prompt‑tuned models. They can be deployed on shared or dedicated hardware, with billing based on usage (tokens or hours).