๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Kartikay Khandelwal
@kakemeister
Realtime AI Inference @ Meta Superintelligence Labs Previously: CTO @WaveFormsAI (acquired by Meta), AI Research @MetaAI, @PyTorch and @Stanford.
๊ฐ€์ž… December 2009
622 ํŒ”๋กœ์ž‰ ์ค‘    1.1K ํŒฌ
Working with @alex_conneau on this has been by far the highlight of my career. Magic happens when you work with the best people. And heโ€™s the absolute GOAT ๐Ÿโค๏ธ
Serving Avatar with real-time videogen models at scale is a very hard AI infra problem. Building an Avatar demo is relatively easy. Maintaining quality and stability while optimizing latency/concurrency enough to make the economics work at scale is much harder. @kakemeister is a Realtime AI Infra wizard ๐Ÿš€
๋” ๋ณด๊ธฐ