企业人工智能：现在、未来与未来（由 Teradata 赞助）.pdf

1、 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Dr.Chris HillmanGlobal AI LeadTeradata Nima BadieyLeader,Strategic AlliancesNVIDIAEnterprise AI:No

2、w,Next,FutureA N T 2 0 2-S 2025 Teradata.All rights reserved.3Improving R&D efficiency with multi-modal AIA large automobile company spent huge amount of time analyzing data related to car field testing.Time-series data from vehicles taking too long to process,is saved in silos and not integrated wi

3、th the text details(test protocols,observations,comments etc)for the test.With Teradata Vantage,we have built a multi-modal AI by combing time-series data with text data Combine embeddings for both time-series data as well as text data into enterprise vector store Used of Agentic AI Natural language

4、 queries,using AWS Bedrock with Claude LLM,on integrated data for faster insights into test resultsValue to customer:Increase in R&D productivity Simplify AI with end-to-end processing within same platform.No isolated vector database point solution needed Leverage Agentic AI to get answers to comple

5、x questions related to vehicle testing using natural language queries 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.R&D Agent in actionLLM This service can answer questions for junior engineer based on past technical document and data(Validation Reports,simulation Results,Design

6、/Control specifications,etc).Question(sample)Which parameter is responsible for the worse fuel consumption Sim results with this specification compared to the previous specification?Why was the battery temperature rise rate higher than expected while running in mode?Why did VSC intervene when it sho