New Delhi, Sept. 7 -- Blue Machines AI has launched Aurora, a speech-to-text model designed for banking, financial services and insurance (BFSI) conversations in India.

The model targets financial conversations where customers switch between English and regional languages, use industry-specific terminology and communicate over noisy or low-bandwidth telephone connections. Blue Machines AI said internal evaluations recorded a 1.51% Semantic Word Error Rate (WER) for English, 2.43% for Hindi BFSI conversations and 5.52% across multilingual speech.

Aurora also recorded a 4.23% BFSI Entity Error Rate for information including monetary amounts, interest rates, policy numbers, account references and transaction IDs.

The model was evaluated a...