Podcast ¦ Punk CX: Douglas Adams’ Babelfish concept just got much closer – Interview with Sharath Keshava Narayana of Sanas

Key Take Aways

• Advances in real-time speech-to-speech translation are rapidly approaching science fiction-level capabilities, with potential implications across financial services for global client communication.
• The invention of breaking down speech into phonetic sounds, rather than intermediate forms like text, enables more natural, emotional, and synchronous translation across languages.
• Traditional translation methodologies—speech to text, intermediate translation, and synthetic output—introduced significant latency and unnaturalness, hindering real-time conversational flow.
• Direct speech-to-speech translation with minimal latency (targeting sub-one second) is feasible and is evolving quickly, supported by advancements in AI and phonetic analysis.
• The technology supports maintaining a user’s natural tonality and emotional nuance, crucial for authentic communication and superior customer rapport.
• A key focus is on solving background noise, dialect, and disfluencies within speech before translation, contributing to higher accuracy.
• Initial deployments have focused on high-impact sectors such as healthcare and customer experience, with applications like multilingual emergency response and improved call centre interactions.
• The platform’s real-time translation can fundamentally improve customer satisfaction (CSAT) metrics and reduce call resolution times in cross-border service scenarios.
• The technology holds significant promise for education, enabling teachers to communicate in native languages, expanding access to quality education across diverse regions.
• The future roadmap includes containerising the translation algorithm for enterprise deployment, ensuring data privacy, and broadening language support.
• Voice, as a human interface, is becoming the dominant communication channel, replacing keyboards, with a strong demand for natural, bias-free, and emotionally authentic communication tools.
• Democratization of technology in recent years allows anyone to build startups with minimal investment, fostering innovation and global digital inclusion.

Key Statistics

• The Sanus app has been downloaded by users from 76 countries within six weeks of launch.
• The platform currently claims a high accuracy rate of 95%+ for core languages such as Spanish, French, German, Mandarin, and Italian.
• Contact time for bilingual conversations initially takes 1 to 3 seconds, with a goal to reduce latency to under 1 second.
• The ratio of shorter business conversations versus longer personal calls indicates diverse use cases, including family, travel, and business communication.
• The speech-to-speech translation market is estimated to be substantial, with significant untapped potential in developing markets and cross-border enterprise communications.

Key Discussion Points

• The limitations of traditional translation methods, especially latency and emotional fidelity, and how phonetic sound analysis overcomes these issues.
• The evolution of speech AI technology from noisy audio cancellation to real-time, emotion-preserving language translation.
• The importance of natural tonality and emotional nuance in authentic communication, especially in customer experience.
• The strategic focus on solving background noise, dialects, and disfluencies for higher accuracy in speech recognition and translation.
• The initial focus on linguistically diverse sectors such as healthcare, emergency response, and customer service to demonstrate real-world impact.
• The potential of instant multilingual communication to revolutionise enterprise and personal interactions.
• Challenges around latency and internet dependency, and ongoing optimisation to achieve sub-one-second response times.
• The phased approach to enterprise integration, starting with cloud deployment and moving towards containerisation for data privacy.
• The application of the technology to enhance customer experience by enabling expert bilingual interactions without interruption.
• The transformative impact on educational access, enabling teachers and students to communicate effortlessly across languages.
• The vision of a future where voice is the primary communication interface, transforming human-machine and human-human interactions.
• The significance of democratised tech development, allowing entrepreneurs worldwide to build and innovate affordably.

Podcast Description

This podcast explores the cutting-edge advancements in real-time speech-to-speech translation technology, with insights from Sharava Nayan, CEO and co-founder of Sanas. It delves into how AI innovations are overcoming traditional limitations of latency, unnaturalness, and emotional fidelity in translation, unlocking new possibilities across sectors such as healthcare, customer experience, and education. The episode examines the strategic implications for global businesses seeking seamless multilingual communication and highlights the broader societal impact of democratized technological innovation. The discussion also touches on future developments, challenges, and the transformative potential of voice as the dominant human interface.