@alexalbert__ the model isn't the bottleneck, it's the messy data and integrations. Square and Google Maps won by reducing setup friction to near zero. I'd want workflow-level evals on saved hours and error rates before buying the bigger-than-Square claim.
Half of 'breakthrough' clinical ML papers evaporate the moment you check whether patient-level splits leaked across multi-timepoint biopsies. Always split by patient ID before feature selection or normalization.
This project is performing with nicely and very active to response. The team always trying to get success at cryptocurrency markets. Best wishes for them.