AI ML Founding Engineer Voice Systems & Full-Stack
VOOK
2 - 5 years
Delhi, Alipur
Posted: 13/06/2026
Job Description
About VOOK
VOOK is building Indias on-device voice memory layer for wearables. We are creating a private, low-power Memory OS that listens, understands, remembers, and acts in natural Hinglish and Indian languages directly on the device where possible.
Our edge is the interaction layer: making earbuds feel like a natural Indian conversation partner with sub-second latency, strong code-switching support, and hardware-rooted privacy. We are early, founder-led, and moving fast with a strong IP focus.
The Role
As a Founding Engineer Voice Systems & Full-Stack, you will own the core real-time voice pipeline and the application layer that turns speech into natural, low-latency conversations. You will work across the full stack (app + backend + voice models), with deep ownership of streaming STT reasoning TTS, turn-taking, barge-in, and the Memory OS.
This is a high-ownership founding role. You will work directly with the founder, make key architecture decisions, and grow into a technical leader as we scale.
What You'll Do
Voice Pipeline & Conversational Systems
Design and ship the real-time voice stack: streaming STT LLM/brain streaming TTS with natural turn-taking, barge-in, and backchanneling.
Integrate, fine-tune (LoRA), quantize, and optimise speech models (Whisper / Conformer / AI4Bharat, VITS / flow-matching TTS, Moshi-style systems) for phone and edge deployment.
Build and improve latency, prosody, and Hinglish / code-switching performance.
Memory OS & Application Layer
Build the capture structure recall memory system and its integration with the voice pipeline.
Design evaluation harnesses and user-correction flywheels to continuously improve quality.
Full-Stack & Product Delivery
Build and maintain the mobile / web app, backend services, and real-time APIs.
Make the system work reliably in noisy Indian environments with graceful offline degradation.
Ship fast, measure real user metrics, and iterate.
Must-Haves
2+ years shipping real products end-to-end (we care more about what youve built than years of experience).
Strong full-stack skills: Python + TypeScript / React / Next.js, backend, APIs, and databases.
Experience integrating and productionising ML / LLM systems (especially speech models).
Comfort with real-time / streaming systems (WebSockets, audio streaming, latency awareness).
High ownership you thrive in ambiguity and can drive features from idea to production with minimal supervision.
Strong Advantage - What Will Make You Stand Out
Hands-on experience with TTS, STT / ASR, voice agents, or real-time conversational systems.
Experience fine-tuning or deploying speech models (especially Indic / Hinglish models from Sarvam, AI4Bharat, or Bhashini).
Knowledge of the modern voice stack: Whisper, Conformer / RNN-T, VITS / F5-TTS / flow-matching, neural codecs, VAD, AEC.
On-device / edge ML experience (quantization, distillation, model optimisation for phones or embedded chips).
Built and shipped a voice-related product or demo end-to-end.
Bonus
Open-source contributions, Hugging Face spaces, or public build logs (especially voice / audio).
Experience with audio DSP or embedded audio systems.
Strong product sense and pragmatic build-vs-buy judgment.
What You'll Get
Founding Engineer title with ESOP real ownership in the company you help build.
Direct collaboration with the founder and a seat at the table on architecture and product decisions.
High autonomy to define engineering processes and culture.
The chance to build category-defining, India-first voice AI hardware + software that can reach millions.
Competitive salary + meaningful equity.
How to Apply
Send the following to careers@vook.in:
Your GitHub / portfolio / relevant links
One project youve built end-to-end (bonus if it involves voice, audio, or real-time systems)
A short note on why youre excited about VOOK and this role
We move fast strong profiles will hear back within a few days.
About the Founder old Ventures:
Ex-TAGG (Audio & Smartwatch Wearables, Acquired by boAt) building VOOK next.
Delhi NCR Region Candidates only.
VOOK is an equal-opportunity team. We hire for what you can build and how you think.
Services you might be interested in
We Search & Apply Jobs for You!
Our team scans through 1000s of opportunities and applies to roles best suited to your profile
Save 100+ hours and focus on what matters - cracking interviews and landing offers.
