scala.bythebay.io: Chalenge Masekera Interview
Recording: scala.bythebay.io: Chalenge Masekera Interview
you my name is challenged mascara I work for salesforce I'm a data scientist for it's called cell suicide Einstein what I do is we're building this very intelligent machine learning pipelines that enable like cells was to provide artificial intelligence to all of these hundreds of thousands of customers so that we can democratize data science so that businesses don't need to employ or look for people with only phd's to benefit from data science so I learnt a scholar when i joined salesforce previously as a death knight is like one of the bigger languages people use a python are so i had a background from python so when i joined sales was like our hostages on scala so that's when I lent how did I learn Scala I think it was a combination of two things that a couple books online which are bought and also i took there's a Coursera class by mattino deskey which I took which is pretty helpful and also I think I have a good number of great teammates you have been used in skala for a bunch of years and they were really helpful with that I think when you're trying to do what you are doing is like build a fool on machine learning stack for a service that's used by hundreds of thousands of customers with like millions of users who are not just like similar but uniquely different in terms of their use cases and like what they're trying to achieve you need one first like a very like highly resilient infrastructure and like in terms of like being able to handle hundreds of thousands of like transactions and also be able to scale quickly and and sort of like handle as new things come up so I think in terms of like if you are into machine learning the one of the bigger things that sort of like famous right now is like spark which is like allows us to like sort of like do distributed computing it with ease we also i think in terms of like data science what you want to move towards is like real-time analytics like when events change you want to be able to predict that so they are a lot of like I stuff like stuff that you are still looking at like a spikes spot streaming link and also like how we can use like cafcass sort of like handle multiple events so in terms of what we are using right now we are using Scala and which can spark and we're also using a bunch of flour scheduler services to sort of like scheduled this job's differently [Music] I'm interested in the AI space and I think ourselves us not the first people to do it but I think so we are looking to I would like to meet with like people also doing sort of like massive scale computing I am sure like the companies are doing that and also like they're a bunch of talks in terms of like deep learning there's a talk at kuntram it's called a deplaning around us I think it will be very interesting to see how they're using Scala or like other like stacks to sort of like dumesh deplaning because it's a very like sort of computing intensive process there's also another talk on what you call it when machine learning architecture and data pipelines I think it's going to be a very good talk which I think I'll learn a lot from then also I think there's another talk I work with a guy but I is very has been is like one of the people talk miss La Scala is going to be talking about like doubt the truth to be a liar you will talk of like what you need to think of in terms of like using Scala to build an automatic machine learning pipeline [Music]