I often use this pattern in my heavy data science work — running scenarios in parallel to compare outcomes. It proved so useful in other projects beyond data science that I decided to formalize it into an API. That’s how Tensorchat was born.