AI ML Founding Engineer Voice Systems & Full-Stack

Role Overview: As a Founding Engineer - Voice Systems & Full-Stack at VOOK, you will be responsible for owning the core real-time voice pipeline and the application layer that transforms speech into natural, low-latency conversations. You will have a high-ownership role working directly with the founder, making key architecture decisions, and evolving into a technical leader as the company scales. Key Responsibilities: - Design and implement the real-time voice stack, including streaming speech-to-text (STT), text generation, and streaming text-to-speech (TTS) with natural turn-taking and barge-in capabilities. - Integrate, fine-tune, quantize, and optimize speech models for deployment on phones and edge devices, focusing on latency, prosody, and performance in Hinglish and code-switching scenarios. - Develop the capture, structure, and recall memory system and integrate it with the voice pipeline, along with designing evaluation mechanisms for continuous quality improvement. - Build and maintain mobile/web applications, backend services, and real-time APIs while ensuring reliable functionality in noisy Indian environments with offline capabilities. - Ship products quickly, measure user metrics, and iterate for continuous improvement. Qualifications Required: - Minimum 2 years of experience in shipping real products end-to-end, with a strong emphasis on practical experience over years of service. - Proficiency in full-stack development using Python, TypeScript, React, Next.js, backend technologies, APIs, and databases. - Experience in integrating and operationalizing machine learning (ML) systems, particularly speech models, and familiarity with real-time/streaming systems like WebSockets and audio streaming. - Demonstrated high ownership, ability to thrive in ambiguity, and drive features from concept to production with minimal supervision. Additional Details: VOOK is an early-stage company with a founder-led approach and a focus on intellectual property. The team is dedicated to building India's on-device voice memory layer for wearables, emphasizing privacy, low-power consumption, and natural language processing in Hinglish and Indian languages. The company's goal is to create a seamless interaction layer that mimics natural Indian conversations, offering sub-second latency, strong code-switching support, and hardware-rooted privacy. (Note: The section on "About the Founder old Ventures" and specific application instructions have been omitted from the job description.) Role Overview: As a Founding Engineer - Voice Systems & Full-Stack at VOOK, you will be responsible for owning the core real-time voice pipeline and the application layer that transforms speech into natural, low-latency conversations. You will have a high-ownership role working directly with the founder, making key architecture decisions, and evolving into a technical leader as the company scales. Key Responsibilities: - Design and implement the real-time voice stack, including streaming speech-to-text (STT), text generation, and streaming text-to-speech (TTS) with natural turn-taking and barge-in capabilities. - Integrate, fine-tune, quantize, and optimize speech models for deployment on phones and edge devices, focusing on latency, prosody, and performance in Hinglish and code-switching scenarios. - Develop the capture, structure, and recall memory system and integrate it with the voice pipeline, along with designing evaluation mechanisms for continuous quality improvement. - Build and maintain mobile/web applications, backend services, and real-time APIs while ensuring reliable functionality in noisy Indian environments with offline capabilities. - Ship products quickly, measure user metrics, and iterate for continuous improvement. Qualifications Required: - Minimum 2 years of experience in shipping real products end-to-end, with a strong emphasis on practical experience over years of service. - Proficiency in full-stack development using Python, TypeScript, React, Next.js, backend technologies, APIs, and databases. - Experience in integrating and operationalizing machine learning (ML) systems, particularly speech models, and familiarity with real-time/streaming systems like WebSockets and audio streaming. - Demonstrated high ownership, ability to thrive in ambiguity, and drive features from concept to production with minimal supervision. Additional Details: VOOK is an early-stage company with a founder-led approach and a focus on intellectual property. The team is dedicated to building India's on-device voice memory layer for wearables, emphasizing privacy, low-power consumption, and natural language processing in Hinglish and Indian languages. The company's goal is to create a seamless interaction layer that mimics natural Indian conversations, offering sub-second latency, strong code-switching support, and hardware-rooted privacy. (Note: The section on "About the Founder old Ventures" and specific application instructions hav