2014—2024

ASAPP Research

ASAPP · Founder & CEO

ASAPP was founded to build a class of product the technology of the day could not yet support: AI that could understand and resolve live customer conversations at the scale of the world's largest service operations. So we decided to build an AI research organization and develop the capabilities to make our product vision come to life. For most of a decade the lab published openly in natural-language processing, speech, dialogue, model efficiency, and reinforcement learning, with work at NeurIPS, ICML, ICLR, ACL, EMNLP, and Interspeech.

Some of the work was fundamental. Some was highly applied. The most interesting projects were both, using the demands of production systems (real traffic, latency budgets, customers who noticed when a model was wrong) to expose scientific questions, then using the answers to make the systems more capable.

Selected work

Full publication list

People

Everything we did was a result of the people we had, and in our case we were able to recruit some amazing researchers. What follows is a selection, by no means a full representation, of the people who joined us and whom I enjoyed working with.

  • Yoav Artzi
  • S.R.K. Branavan
  • Howard Chen
  • Hui Dai
  • Ethan Elenberg
  • Varun Gangal
  • Kyu Han
  • Shawn Henry
  • Arzoo Katiyar
  • Kwangyoun Kim
  • Barrett Lattimer
  • Tao Lei
  • Tao Ma
  • Ryan McDonald
  • Geoff Pleiss
  • Ramya Ramakrishnan
  • Suwon Shon
  • Paloma Sodhi
  • Prashant Sridhar
  • Kilian Weinberger
  • Jeremy Wohlwend
  • Felix Wu
  • Yi Yang
  • Lili Yu