Deep in the Heart of Texas - How Indeed Embraced Deep Learning to Help People Get Jobs!
Over the past decade, Deep Learning has demonstrated significant performance gains in numerous important application domains. This has resulted in a marked shift in the machine learning landscape. Although we dabbled prior, in 2017 Indeed began concerted efforts to explore how these techniques could be leveraged to deliver transformational business value. Ultimately, we decided to hire a technical Deep Learning lead to accelerate these initiatives. In this talk, we will outline our deep learning journey. We’ll describe the progress we’ve made so far, the challenges we’ve overcome, and the lessons we’ve learned along the way. We’re hoping you can learn from our experience, both the successes and the bruises we’ve amassed. Along the way, we’ll discuss the Deep Learning solutions we’ve deployed - including state of the art natural language processing techniques, deep learning based recommender systems, and methods to improve existing systems using representation learning.
Thom Lake is the Deep Learning Technical Lead at Indeed. There, he finds ways to use deep learning to "help people get jobs," and guides the application of modern machine learning techniques across the company. Throughout his career, he's had the opportunity to design and develop machine learning solutions for a variety of domains, from motion recognition in resource-constrained embedded environments, to product recommendation on one of the largest e-commerce websites in the world. His interests include neural networks and deep learning, inductive bias, natural language processing, and the interaction between humans and adaptive systems.