Models Product Manager | Sarvam AI | Bengaluru

Role overview

Sarvam AI is hiring a Product Manager (Models) to lead product charters across core model capabilities, including Automatic Speech Recognition (ASR), Text-to-Speech (TTS), and model evaluation pipelines. The role is based on-site in Bengaluru, Karnataka, India, and sits at the intersection of product, research, and engineering.

Key details

• Job title: Product Manager (Models)

• Company: Sarvam AI

• Location: Bengaluru, Karnataka, India (On-site)

• Function: Product Management (AI / ML models)

• Employment type: Full-time

What you’ll do in this role

As Product Manager (Models), you will own the strategy, roadmap, and success metrics for Sarvam AI’s core model capabilities:

• Define and drive product roadmaps for ASR, TTS, and model evaluation systems, aligned with user and business needs.

• Partner with research and engineering teams on data requirements, training cycles, and model release planning.

• Design and own evaluation frameworks and quality metrics (e.g., accuracy, latency, MOS, robustness) to track model performance.

• Run structured experiments (offline tests, A/B tests, human evaluations) to quantify improvements and inform priorities.

• Work with infra and data teams to ensure scalable deployment, monitoring, and reliability of models in production.

• Translate high-level user and business requirements into clear, measurable model goals and milestones.

• Communicate priorities, timelines, risks, and trade-offs clearly to stakeholders across product, research, and engineering.

What you’ll bring to the role

Sarvam AI is looking for a product manager who is deeply comfortable with applied AI and model lifecycles:

• 3+ years of experience building AI / applied AI / ML products or managing model training pipelines.

• Strong understanding of the end-to-end model lifecycle – training, tuning, deployment, and monitoring.

• Excellent analytical and data-driven decision-making skills; comfortable working with experiments and metrics.

• Ability to collaborate closely with research scientists, ML engineers, and product teams.

• Exposure to speech technologies (ASR, TTS) or prior work on evaluation pipelines is a strong plus.

• Familiarity with quality and performance metrics such as WER/CER, MOS, latency, and robustness measures.

• Hands-on comfort with data and tools (e.g., SQL, Python, ML evaluation toolchains) is preferred.

What to expect

This role places you at the heart of Sarvam AI’s model strategy, working on high-impact ASR and TTS systems and the evaluation pipelines that determine product quality. You will help shape how models are trained, evaluated, launched, and iterated, with direct influence on user experience and business outcomes.

How to apply

To apply, visit the official Sarvam AI job page and submit your application (you can autofill using a .pdf or .docx resume):

• Official listing & application:
  https://careers.kula.ai/sarvam-ai/13082

GrowInJob.com does not accept applications or CVs for this role and does not influence the hiring process. Please apply only via the official Sarvam AI / Kula careers link above.