Staff Engineer @ Uber 🚀 Open source work LLM Coder Browser-based multimodal AI that runs locally Audio, text and image inputs Custom fine-tuned LLama 3.2 Vision models Open source repo 🎯 Qwen2-VL on the browser Achieved first-ever browser deployment of Qwen2-VL via ONNX conversion. Available on my Huggingface profile 💼 Impact at Uber Edge caching initiative: ⚡️ 50% reduction in TTFB 🌐 Serving 1b+ requests a month Candidate offer tools revamp: 📈 20 day reduction in time to offer 👥 Used for 1M+ candidates AI test coverage: 🎯 Coverage increased by 50% 🔄 90% automated refactoring