Decision makers across the global finance industry look to Bloomberg to deliver timely political, business and financial information and insight. Today, a large part of this insight is derived using artificial intelligence (AI). The practice of AI at Bloomberg presents its researchers with the unique challenges associated with building low-latency, high-precision, and highly-available systems that need to find needles in haystacks of data. In this talk, we will introduce the wide variety of problems that we solve using AI at Bloomberg. We will then lead an in-depth discussion of our Twitter processing pipeline, which has enabled investors to sift through the daily stream of tweets in order to discover financiallyrelevant content. Finally, we will provide a brief overview of the cutting-edge AI research that our teams have recently published.