ABSTRACT: Applications such as Internet telephony (VoIP) and online gaming place stringent demands on network performance in terms of latency, jitter, and packet loss. The best-effort Internet often falls short of meeting these requirements. In the PinDrop project at Microsoft Research, we are developing the networking substrate for supporting high-quality real-time streaming. Our approach centers on three pillars: (1) leveraging data-driven insights, (2) exploiting network path redundancy, and (3) performing wireless-specific optimizations. After touching briefly on these, I will focus on Kwikr, which uses WiFi hints to improve bandwidth estimation for Skype calls. Kwikr uses a simple technique called ping-pair to estimate and attribute the queuing delay at the WiFi downlink, which enables more informed bandwidth adaptation than the state of the art. BIO: Venkat Padmanabhan is a Principal Researcher at Microsoft Research India, where he founded the Mobility, Networks, and Systems group in 2007. He was previously with Microsoft Research Redmond, USA for nearly 9 years, and holds a B.Tech. from IIT Delhi and an M.S. and a Ph.D. from UC Berkeley, all in Computer Science. Venkat’s research interests are broadly in networked and mobile systems. He recently received the Shanti Swarup Bhatnagar Prize and the inaugural ACM SIGMOBILE Test-of-Time paper award. He has been elected a Fellow of the Indian National Academy of Engineering (INAE), the IEEE, and the ACM. He presently serves on the SIGCOMM Industrial Liaison Board. He can be reached online athttp://research.microsoft.com/~padmanab/.