● Object detection using a dataset of 5,000 annotated images. ● Processing a diverse dataset of over 10,000 audio clips (2 to 4 secs), converting these into vector embedding to train robust representation models. ● Search and match capabilities