In this presentation we will share some of our experiences in building and deploying cutting edge NLP technologies at scale for enterprises. We will focus on domain specific information retrieval and question answering systems, and discuss the challenges, constraints, and opportunities around creating real-world enterprise-grade experiences. We will cover a number of areas including flexible content ingestion from variety of sources and formats, efficient high fidelity retrieval, and trustworthy response generation. We will also discuss data collection challenges for model training and evaluation, security, and user level access control, and their crucial role in realizing high impact results. Many of these challenges, particularly around trustworthiness and security, have implications that extend beyond enterprise applications, and their solutions will ultimately influence the broader adoption and practice of artificial intelligence.