When I saw what bitdrift was doing I had flashbacks to moments when I wondered how it was possible to store so many metrics and generate so many logs, yet still never have enough context when something went wrong. Exciting to see how bitdrift's approach will impact mobile observability, and hopefully observability as a whole! It would be really fascinating to hear about the conversations that lead to the decision that it would be worth it to build bitdrift instead of trying to piece something together with existing solutions.