Moshi processes 300 data points per second for traders
Moshi provides real-time insights for financial traders by processing 300 new data items every four seconds. The platform integrates structured and unstructured data to generate portfolio-specific recommendations.
Moshi, a startup founded by former JP Morgan derivatives trader Anika, provides real-time insights for financial traders in commodity, energy, and macro markets. The platform processes data at a volume of 300 new items every four seconds. It integrates structured data, such as futures time series and technical indicators, with unstructured sources including news, X (Twitter) data, and podcast audio. Moshi also ingests private user data, such as internal LNG ship counts through the Strait of Hormuz and specific analyst subscriptions.
The system uses a vision-language model to process messy PDF uploads and constructs an n-dimensional graph to generate portfolio-specific recommendations. Anika, who holds a degree from Warwick and a Fulbright scholarship, previously published research at ICLR with Stability AI and collaborated with NASA. The company describes its product as a hybrid of Bloomberg, Claude, and Netflix, emphasizing a recommendation engine that surfaces second-order market impacts rather than just news streams. In the same amount of time, Moshi has learned 300 new things, according to the company.