Shopee's Custom LLM Scales 113x in 8 Months, Cuts Fraud Costs by 90%
nvidia · x · 2026-08-21
Shopee published results on Compass, its frontier LLM built for multilingual e-commerce in Southeast Asia. To bridge the "jagged frontier" where general models fail at local tasks, Shopee used the NVIDIA AI Factory stack. Key outcomes include API token scaling from 3 billion to 340 billion monthly (113x growth) and a 50x speedup in fraud detection at 90% lower processing cost via TensorRT-LLM. The tech stack utilizes NVIDIA Megatron-Core for pretraining, NeMo for post-training, and runs on A100, H100, and Blackwell GPUs.
More from Companies & People
- Accenture caps Claude reimbursement at $29/month, employees used to $1,000 scramble — _rchaves_ · 2026-10-16
- AI billboards take to the LA skies as startup hires plane to tow ad banner — TarlonKhoubyari · 2026-10-16
- SUNY Albany to offer Topological Deep Learning course, from Perceptrons to WL test — burny_tech · 2026-10-16
- Palantir opens third year of merit-only Meritocracy Fellowship applications — eliano · 2026-10-16
- Why hiring 'technical people who can create content' is so hard, from a Cerebras dev — charles_irl · 2026-10-16
- Sentry to host System Update panel on agentic coding with OpenAI, Linear, DoorDash leaders — zeeg · 2026-10-16