This line "Set up low-latency, elastic, highly-available, and cost-effective inference close to your data." is the problem I've found that annoys everyone - Sof

This line "Set up low-latency, elastic, highly-available, and cost-effective inference close to your data." is the problem I've found that annoys everyone - Software Eng don't understand MLOps, Data Scientists don't understand systems programming, and everything ends up costing way too much money and taking way too much time. Even with a magic API the latency still isnt good enough, so the choice is often entire ML solutions like the OPs product, months of development time, or a really expensive container. Shameless plug since my Show HN went completely ignored earlier today ;) ...but that's why i built this: https://news.ycombinator.com/item?id=30428664 https://news.ycombinator.com/item?id=30428664