My ongoing research is around building most effective datasets, recipes and data flywheel for language models to reveal their potentials further. Sitting at the intersection of Gemini and Bard, I improve language model’s core capabilities and horizontal quality with whatever I can do and whatever I can learn to do.