Stop Generative AI Envy with RAG (No Data Scientists Needed!)
Transcript
[Music] do you have generative AI Envy you're looking at peers and other parts of your area of expertise you can be a marketer you can be a financial services advisor and they're using generative AI to become more efficient they're using it to create new products they're using it to get more data or more insights to their customers and content consumers than you can because you work in an industry that frankly you can't take the 10 years of customer artifacts that you have and just simply upload
it into a public repository where a large language model could learn from it but you think you know I don't have the data scientist Keith I don't have the data scientist that can take all of my data training on mod models and get these insights that you know that's within your data this is where retrieval augmented generation rag comes into play you can take a typical Enterprise server with some gpus a hefty box run one of these large models such as granite or llama on that
act on that server upload your data and this is the key part what does it mean to upload your data to a model do you need to have these data scientists retrain the models and do this night after night this is where rag comes into play you can just simply take the flat files put them into a folder run a batch process that connects the data set set to your large language model you get much more accurate results based on your data you can ask a
question such as give me the top 10 customers by region based on a criteria that you set give me insights based on whatever insight you're looking for data criteria you're looking for all based on the different models that you select the complete flexibility without hiring data scientists you want to put this to test you want to learn more visit Dale on the web dell.com want to learn more about our research at the futurum group futurum group.com where we talk about this quite a bit from a
wide industry perspective you want to find out more about me I'm at CTO advisor talk to your next CTO dose [Music]
