Senior AI Engineer(TTS)

Department Icon Data Science Analytics & Machine Learning
Posted: 1 month ago
8-10 years
Bengaluru / Bangalore, Karnataka
Data Science Analytics & Machine Learning
work from office
Full Time
Closes 20 Nov 2026

Posted: 1 month ago
Let Jobaaj One apply for you

An AI agent that applies to roles like this on employers' own sites. 25 applications free, no card needed.

Job Description
About Company
Similar Jobs
Please verify your account first! Send OTP

Job Description

Senior AI Engineer, Text-to-Speech (TTS)

www.gnani.ai

Location: Bengaluru Type: Full-time(work from Office) Function: AI Team

About the role

While youre reading this, Gnani is talking to customers across India — in Hindi, Tamil, Telugu, and dozens more languages, switching between them mid-sentence without missing a beat. That voice is our TTS stack. Its the difference between an AI that sounds robotic and one that sounds human, expressive, and unmistakably Indian.

Were building the most natural voices in Indian languages, at production latency and enterprise scale — and we want you to lead the research behind them. As Senior Staff Engineer for TTS, you own the stack end-to-end: advancing quality and expressiveness, engineering it for real-time performance, and building commercial-grade voice cloning. This is a senior individual-contributor role: you set technical direction, design experiments, and mentor the researchers around you — without a people-management load.

Core mandate

Research

Advance TTS quality, expressiveness, and multilingual coverage — experiment across architectures

Production & performance

Take models to production with low latency, high throughput, and reliability at scale

Voice cloning & customization

Build commercial-grade cloning and voice design for media and enterprise use

Technical leadership

Guide juniors, design rigorous experiments, and own the standard for how TTS gets built here

What youll drive

Research & modeling

  • Own the TTS modeling roadmap — architecture selection, quality, expressiveness, and multilingual / code-switched coverage across Indian languages
  • Experiment across TTS architectures (autoregressive / LM-based, non-autoregressive, diffusion / flow-matching, neural-codec–based) and translate findings into shippable systems
  • Push on the hard problems: prosody and emotion control, naturalness, code-switching (e.g. Hinglish), G2P for Indic scripts, voice design, audio tags, and speaker / style consistency
  • Define and track quality evaluation — MOS / CMOS, intelligibility (WER via ASR), speaker similarity — and make architecture calls grounded in data
  • Participate in benchmarks and publish the work in international conferences

Production & performance engineering

  • Design TTS systems for production: streaming synthesis, low first-audio latency, favourable RTF, and high concurrency on GPU
  • Own the performance envelope — batching, quantization, kernel / inference optimization, and cost-per-hour of audio
  • Optimize production inference using ONNX, TensorRT, NVIDIA Triton, quantization, batching, and GPU-efficient serving architectures
  • Partner with platform and infra teams on serving, scaling, and reliability — latency SLAs, uptime, error budgets
  • Build evaluation and regression harnesses so quality and latency dont silently regress release to release

Voice cloning & customer quality

  • Build and improve commercial-grade voice cloning for media and enterprise applications — few-shot / zero-shot cloning, voice design from a brief, and controllable style
  • Work directly with customers to resolve production issues, fine-tune voice quality, and address cloning fidelity, consent, and safety concerns
  • Turn recurring customer quality issues into systematic model and pipeline improvements

Technical leadership & research presence

  • Technically guide junior researchers and engineers — experiment design, tracking, code and research review, and career-shaping mentorship
  • Set the bar for rigor: reproducible experiments, clean baselines, honest evaluation, and clear write-ups
  • Stay current with TTS research and bring the relevant frontier into Gnanis roadmap
  • Publish and represent Gnanis TTS work externally — international conferences, talks, and community engagement

Who you are

EXPERIENCE WERE LOOKING FOR

  • 8–10 years in speech / ML research or engineering, with substantial contribution to production-grade TTS systems
  • A track record of taking TTS from research to production with a focus on latency, quality, and performance at scale
  • Hands-on experience across multiple TTS architectures — youve built, compared, and shipped, not just read papers
  • Experience building commercial-grade voice cloning, ideally for media applications
  • Good publications at international conferences (e.g. Interspeech, ICASSP, NeurIPS, ICML, ACL) in speech / audio / ML
  • Strong system-design instincts for the engineering challenges of productionizing speech models

WHAT MAKES A STANDOUT CANDIDATE

  • Deep familiarity with the Indian voice landscape — Indic languages, code-switching, dialect and accent diversity
  • Demonstrated ability to guide juniors and lead experiment design and tracking across a team
  • Up-to-date with TTS research, and able to separate durable advances from hype
  • Experience with neural codecs / tokenizers (RVQ and similar) and LLM-style TTS
  • Comfortable operating with high ownership in a fast-moving, post–Series B environment

TECHNICAL FLUENCY — A MUST-HAVE

  • Deep understanding of modern TTS: acoustic modeling, neural vocoders and codecs, autoregressive vs non-autoregressive synthesis, diffusion / flow approaches, LLM / codec-token TTS, and prosody / duration / pitch control
  • Fluency with TTS evaluation — MOS / CMOS, WER (intelligibility via ASR), RTF and latency, speaker similarity for cloning — and what drives improvements or regressions on each
  • Working knowledge of production ML systems: streaming inference, GPU serving, batching, quantization, and the latency / quality / cost trade-offs that decide what ships
  • Judgment on evaluation and benchmark design: good test sets, fair comparison across languages / accents / domains, and not overfitting to a benchmark

Why Gnani, why now

Gnani.ai is one of the few companies in India doing original speech research at depth — our models are trained on real Indian language data at scale, not adapted from English-first architectures. As Senior Staff Engineer for TTS, youll own one of the most technically demanding surfaces in Indian voice AI, with the freedom to set direction and the impact of seeing your models power real conversational products across banking, healthcare, logistics, and government services.

High visibility, high impact, and a clear path to the most senior technical roles as the company scales.

Skills

Ai EngineerAiMlLlm

If a job posting appears fraudulent, asks for payment, contains misleading information, or violates our guidelines, please report it immediately. Our team will review it promptly, Jobaaj does not charge any fee from the applicants.

About Company

Gnani.ai is Indias leading enterprise voice AI company, building the language infrastructure that powers intelligent voice agents, speech recognition, and text-to-speech systems at scale across 40+ languages. Our models are deployed across government, BFSI, telecom, and enterprise verticals, processing over 30 million voice AI calls a day. Founded in 2016 and backed by Samsung Ventures and Info Edge Ventures, we are one of four companies selected under the IndiaAI Mission to build foundational AI models for India.

Read More

Similar Jobs

View all
Loading...

Looking for something similar?

Bag Logo
Jobaaj
Get jobs like this by email

New Data Science Analytics & Machine Learning roles in Bengaluru / Bangalore.

Job Alert
Google hiring for Specific Roles Apply Now!
1 min ago
New Opportunity
Amazon is hiring freshers Apply Now!
5 min ago
Featured Jobs
Microsoft opening 50+ positions Apply Now!
10 min ago

One email a day. Unsubscribe any time.