Democratizing Analytics at Micron Technology
Ionicia Dembi is Senior Manager in Smart Manufacturing and AI at Micron Technology. For years, she has supported the organization’s yield enhancement, product yield engineering, quality and data analytics. As a JMP user, Dembi and her teams utilize JMP as one of their analytics tools for statistics, modeling and data visualization.
Transcript
00:00:00 At Micron we're very data driven company. There is expectation for people in all sorts of roles that they need to learn some components of analytics or statistics because data scientists are not enough [Music] We have so much data in our manufacturing process that you can have thousands of data scientists and you will still need to learn new things. We need domain experts to partner very closely with the data scientists. You don't necessarily want process Engineers to become experts in Python or R some of these other open source languages that are out there. JMP helps guide them to analyze things and try different ways to analyze the data different ways to visualize the data they do the analysis they see their results they get excited you can see that they want to learn more they trust the data or they understand the data more they understand some of the limitations and they can help guide and work with the data scientists for future
00:01:08 work. That does provide a lot of value and our leadership is on board and they're encouraging people to learn and to grow. The more we talk about it analytics and more people understand what it is that it's not scary it is just a tool is just something that you learn and with JMP a big part in the UI component is having the visualization of the data. There are some people that you show them the tables or you show them summary statistics and they're like they don't quite understand it. With the Graph Builder capabilities and the visualization of the data people that were maybe not very advanced in statistics see the data visually and so there was a that connection. I can see it now rather than just tables or statistics. At Micron there are training tracks that people are in based on what their position is at the company and then as they move or progress through their careers. It starts basically with the technicians and then it goes all the way up to the
00:02:12 senior leaders. We have basic statistics courses and we are using STIPS as part of the statistical training. We have Coursera licenses that people can go in and do more advanced training and uh I think modularizing it helps a lot. We're fortunate that we have this partnership with JMP that if there is an advanced topic that maybe nobody within the company knows we can partner with JMP to say hey can we get training tailored to us because we have those super users that really want to know these topics and they want to know how they can do this type of analysis. We also have those people that love JMP and anytime something new comes out they want to try it they want to improve the process and improve yield quality and so on. If they have somebody within their team that is a JMP power users they may ask them to build add-ins so that way the rest of the team will be able to do some of that work without having to know everything and actually once you have
00:03:12 the add-ins a lot more people can do the analysis and you know that they do it consistently the same way and get the same results. We've done quite a bit working with JMP providing feedback what our users are saying what are the things that they would want to have as they're doing analytics and people were very excited and happy that JMP listened to us and I'm sure other people in the JMP user Community being able to query the databases directly from JMP. So previously we would have people that would build add-ins that non-JMP users or JMP users that were basic users would not be able to do but then the basic users would still have to go in and extract the data and then try to figure out different ways to extract the data before they would be able to do the analysis utilizing the add-ins. So once we had the capability of querying the databases then the JMP power users could just create those add-ins extracting the data and doing the analysis that
00:04:13 was a you know a very good thing and again feedback to JMP and JMP listening to us all of our businesses they're just becoming more and more complex we get more and more data more and more different types of data and so we do need to have everybody rise and increase their skill sets. Within the business at Micron we don't have to convince people. Everybody has seen the value on analytics and that it helps us analyze things, process improvements, yield Improvement, quality improvements, time to market that it helps the business all of that is huge and success breeds success basically. Micron leadership is very focused and I want us to be able to make decisions that are data driven and they see their results and who they get excited about it. [Music]