1、AI for BI Without the BSJamie Davidson,Co-Founder&President OmniJune 2025Is AI the holy grail of self-service weve been waiting for?2LLMs are smart,but they lack context.3How many users do we have?“Is that different from active users?“Does it include employees?Partners?“AI alone cant solve the hard
2、parts of BI.45Data quality mattersGuardrails for security&governance Defined,shared vocabulary for humans&AIHuman-in-the-loop to check,rinse,&repeatKey considerations for useful,reliable AI6Semantic layer(n)Semantic layers define business logic generally sitting between the data store and business u
3、sers7Semantic layers create a shared vocabulary for humans&AISemantic model8Semantic layers enable human-in-the-loop workflowsSemantic modelImprove accuracyAI workflowsSemantic layers enable human-in-the-loop workflowsSemantic modelImprove accuracyAI workflowsSemantic layers enable human-in-the-loop
4、 workflowsSemantic modelImprove accuracyAI workflowsSemantic layers enable human-in-the-loop workflowsSemantic modelImprove accuracyAI workflowsSemantic layers enable human-in-the-loop workflowsSemantic modelImprove accuracyAI workflowsCurate&editSemantic layers enable human-in-the-loop workflows14B
5、usiness usersEnable self-serviceData teamReduce manual workCustomersLevel up their experienceCreate tailored AI experiences to meet desired outcomes15Level up your customer experienceEmbedded AI analytics&custom AI data productsEnable customers to:Get started with analyses regardless of their skills
6、etEnable customers to:Get started with analyses regardless of their skillsetEnable customers to:Get started with analyses regardless of their skillsetEnable customers to:Ask follow up questions with help from AI20Enable business users to self-serve dataQuery,summarize and visualize data with AIEnabl