Moshi builds recommendation engine for commodity traders

The system builds an n‑dimensional graph from cleaned and tagged data, then pushes tailored alerts to users, offering second‑order connections that a trader might miss—for example, how an energy‑equities move could affect oil prices. Moshi also provides confidence scores for conflicting sources and uses a vision‑language model to extract information from PDFs. The company is handling inbound interest from traders and AI engineers, scheduling meetings and a webinar to manage demand.