Hey HN. I have been building TESSRAL for a while. It is a deeptech product which lets you train your own image or voice AI models with a small amount of processing power and a small dataset belonging to you. Or you may use other TESSRAL users datasets if they let you. The point is we don t need huge datacenters in order to train AI models. No massive GPU clusters, no renting compute from someone else s cloud.Just your own hardware. If you have any questions (the technology, the models etc.) I would be happy to answer them.