@arshadhamna1829929

Chaukidar helps teams evaluate whether AI models remain safe beyond English-only testing. It audits LLM responses across South Asian languages and harm categories such as fraud, cyber abuse, self-harm, hate, and violence. The platform compares English seed prompts, translation-baseline prompts, and native-adapted prompts to reveal where model safety behavior changes across language, culture, and phrasing. Built for AMD Hackathon Track 3, Chaukidar uses AMD ROCm/vLLM notebook results for compute-backed audits, supports Fireworks model comparisons, re-judges imported outputs with a multilingual judge, and generates dashboards, risk heatmaps, track comparisons, and audit reports.

To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits. All hackathon participants will receive on hackathon launch additional compute and API credits for their projects. Ready to build what's next on AMD? Sign up with AMD below to secure your spot.