Predii introduces Generative AI ‘Co-Pilot’ for repair shops

Predii 360 responds to questions and produces intelligence to enhance productivity of service writers, techs, parts staff and more

Palo Alto, Calif.—Aftermarket AI innovation leader Predii Inc. has introduced the industry’s first automotive AI “co-pilot” designed to augment and support automotive service and repair workflows.

Predii 360 responds to automotive service and repair questions and produces actionable intelligence to enhance the efficiency and productivity of service writers, technicians, parts counter people, and other industry professionals. 

Predii executives will demonstrate Predii 360 during the AAPEX 2023 Show, Oct. 31-Nov. 2, in Las Vegas.

Predii 360 is designed to integrate optimally with custom service and repair data. Models are hosted in an isolated environment, protecting intellectual property, and enabling companies to innovate and commercialize new solutions using Generative AI.

“The future of the automotive industry is very quickly moving toward one-point-entry AI solutions that allow users to move beyond siloed information and interact with systems in a human-like manner,” said Predii Founder and CEO Tilak Kasturi. “Having an AI collaborator in the loop in expert-knowledge-dependent industries will become the norm very soon.”

Predii 360 leverages multiple components within the company’s technology portfolio:

• Predii AI Pipeline, which extracts features from billions of automotive service data points.

• Predii Intelligent Intent Response Network (IIRN), which understands industry specific intents of questions and responds by pulling in custom knowledge sources.

•  LLM, an automotive specific, fine-tuned Large Language Model that allows for use-case specific knowledge ingestion and retrieval framework.

Predii 360 responses are evaluated through Okareo, a product for teams using LLMs to detect and measure response accuracy and predictability.

A public demo of Predii 360 is available at https://predii360.ai. This early beta version has access to limited data sets such as National Highway Traffic Safety Administration public complaints and recalls, technical service bulletins and vehicle service manuals, and leverages parts catalog API integration.