About Us
Technology-driven, scenario-focused
Who we are
NGJOO AI is an AI team based in Guangzhou, China, focused on core LLM technology: model training and fine-tuning, AIGC application development, and AI red-team security evaluation.
Our team members come from front-line AI research and engineering, covering the full stack from data processing and distributed training to model deployment. We compete in AI algorithm contests under the team name “拒做韭菜”.
Selected contest awards
- CCKS 2026 Task 5 (Text Attribute Extraction under Complex Constraints) — 1st place
- Tianchi Better Synth Multimodal LLM Data Synthesis Challenge — Champion
- BetterMixture Large-Scale Data Mixture Challenge — Runner-up
- IJCAI 2025 Deepfake Detection Challenge, LLM Safety Track — Excellence Award
- CCL2024-Eval Chinese Frame Semantic Parsing — 3rd place
- OpenAtom Open Source Competition · Intel Enterprise AI Solutions — 3rd prize
- openKylin AI Computing-Scheduling Platform Challenge — 3rd prize
- vLLM & LLaMA-Factory Performance Optimization Challenge — 3rd prize
What this site does
- AI Trending: daily tracking of trending GitHub open-source projects with Chinese and English analyses.
- Blog: hands-on experience in LLM training, inference optimization, and AI safety.
- Paper deep-dives: engineering-oriented readings of top-conference papers.
- Products: self-built tools around the LLM toolchain.
How we produce and review content: see our Editorial Policy.