Scale By the Bay 2023: Interview with Riley Hun.
Recording: Scale By the Bay 2023: Interview with Riley Hun.
so my name is Riley hun I'm uh working at Autodesk as a ml platform engineer I'm part of the auto machine learning platform team and we're building a centralized ml platform at aut um to serve many of our customers and um building an acceler their model developments from research to production so um mostly the research scientists um there are various different research teams at Autodesk and they're using AI for different um use cases um some more traditional ML and then others doing more Cutting Edge stuff um specifically in the 2D and 3D geometric design space and then we are on more on the infrastructure side so we're basically enabling those research teams to to move away from their bespoke um uh ml uh platforms and then using our more centralized uh ml platform and then we're trying to evolve that into an AI platform so um we're basically doing a lot of um looking into provisioning different tooling um in the ML and AI space um basically training their training the train providing them the tools to train their models and then um looking into feature stores and um data ETL data pipelines and then serving their models in in production as real-time API endpoints um uh right now our research team is um working on a LGE like complex large scale uh model and that's as much as I can divulge but um we are working with them to uh move that model into production so that's probably the main the main project and then next year I'll be focusing on um distributed uh inference um which is definitely like of Paramount importance in our organization because obviously llms and um large scale models are all their AG nowadays so we need to uh be sure that we have our ml AI platform is well equipped to handle um handle that kind of use case and then also we're trying to make um training and fine-tuning open source llms more readily available and um for our different stakeholder teams as well so um definitely the talk that looks into um in integrating uh rag um for for llms definitely we could use that talk to learn about some of their experiences because we're probably going to have to do something similar for our AI platform so I'm sure we can learn a lot from that talk um rag is going to play a huge huge part in um making ouro platform and AI platform more um readily Avail uh readily equipped to handle um L LMS and generative AI so we're we want to take advantage of what we can learn from from that talk to to do that perfect so I would definitely recommend um my friends to check this venue out um it there's a lot of diverse um different like different talks with uh that delve into many different top topics um and you're likely to find um a talk that that is interesting to you that and perhaps that can uh help you um apply apply what you learn into like a more to to something more practical in in your the in your day job or your work um and it this um venue definitely embodies like collab a collaborative Spirit um everyone is very supportive and you get a good Community Vibe um from scale by the Bay