As a data scientist, I specialize in extracting actionable insights from mobility and spatial data. My expertise includes developing low-cost, no-training solutions by combining hybrid models like YOLO and ViT embeddings. Leveraging synthetic data generation in Blender, I have significantly reduced company resources spent on data collection and labelling. I have also contributed to projects that detect deep fakes generated by models such as SD and GAN using innovative solutions like LNPs and Fourier transform.In addition, I have deployed the Whisper model on Edge devices to transcribe audio without requiring Internet access. My work with LLM Blender (PairRM) to compare GPT-4 and Mistral responses yielded exciting results. I have trained a GPT-2 model to provide directions using location information.I am passionate about transforming data into insights and continuously strive for perfection in my work. My goal is to leverage cutting-edge technologies to drive impactful results.