Benefits and Challenges of combining Deep Learning and Search
Deep Learning has fundamentally changed Information Retrieval. Traditional search retrieves text documents based on text queries. Today, however, companies want to search for images, videos, customers, jobs, shopping catalog items, friends, places, and many more. For such searches, Deep Learning models provide the most accurate results. Unfortunately, deploying and serving machine learning models at massive scale is still a herculean effort even for highly tech-heavy companies. This talk will survey some of those challenges. We will also introduce a new cloud-managed service by HyperCube which serves such real-time workloads.