Back
YC backing, metro location, and a mid-level ML hiring range increase applicant competition significantly.
Role requires specialized speech and production ML expertise, so industry/domain fit is highly important.
Explicit 3+ years plus mandatory PyTorch, distributed training, and production ML skills make filters strict.
Login to See Your Match Score
Create a free account or log in to unlock your CV match score across:
Own the end-to-end lifecycle of building, evaluating, deploying, and improving ML models powering millions of voice AI conversations.
Design and build data pipelines handling conversational voice data across Indian languages, accents, and telephony conditions.
Deploy models in latency-sensitive production, monitor performance, debug issues, and iterate rapidly to improve quality and reliability.
3+ years of hands-on machine learning experience with real-world training of models.
Strong Python and PyTorch skills including distributed training and modern fine-tuning techniques (e.g., LoRA, QLoRA, DPO, RLHF).
Experience designing and implementing data pipelines for collection, cleaning, labeling, augmentation with a focus on data quality.
Not explicitly mentioned: formal degree requirements, notice period, or strict location constraints.
Experienced in shipping production ML models with a bias toward delivering working models over perfect research prototypes.
Strong evaluation discipline, including building benchmarks and human-in-the-loop pipelines for quality measurement beyond demos.
Prior experience with speech, real-time inference, voice AI domains, or multilingual Indian language models is a strong advantage.