1、Fidelity and AWS are two separate companies.Unless otherwise noted,the opinions provided are those of the speaker,not necessarily those of Fidelity Investments or its affiliates.I N D 3 3 2 3Fidelity Investments:Text-to-SQL for data analytics at enterprise scaleRay ZhangVP,ML Engineering Practice Le
2、adFidelity InvestmentsManish WorlikarHead of FI Artificial Intelligence CoEFidelity InvestmentsVikram Elango Sr.AIML Specialist Solution ArchitectAmazon Web ServicesFidelity and AWS are two separate companies.Unless otherwise noted,the opinions provided are those of the speaker,not necessarily those
3、 of Fidelity Investments or its affiliates.2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.AgendaThe journey to enterprise-scale text-to-SQLBuilding the foundationdata and semanticsEnabling intelligent queryingMeasuring,scaling,and learning 2025,Amazon Web Services,Inc.or its affi
4、liates.All rights reserved.The journey to enterprise-scale text-to-SQL 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.How did this start?User-centered observationWe observed daily tools and navigation patterns to understand how business users operateInformation access challengesI
5、nformation was fragmented,causing delays and frequent tool switchingInsightful discoveryUsers asked smart questions,but answers often required deep searches or outside help 2025,Amazon Web Services,Inc.or its affiliates.All rights reserved.Enterprise data analyticsChallengesComplex data,slow insight
6、sLarge,fragmented datasets make it hard to extract timely,meaningful insightsslowing decision-making and innovationTraditional SQL:a bottleneckManual SQL queries require technical skill and deep data knowledge,limiting access to specialistsTechnical barriers to accessSQLs complexity creates dependen