Job Description

About Deepgram: Deepgram is the market-defining technology platform underpinning the emerging trillion-dollar Voice AI economy. We engineer production-grade cloud APIs and self-hosted foundation models for ultra-low latency speech-to-text (STT), text-to-speech (TTS), and advanced voice agents at scale. Backed by $215M in venture funding from premier investors including NVIDIA and Y Combinator, Deepgram has processed over 50,000 years of audio to empower more than 200,000 developers worldwide.

Position Overview

We are seeking a seasoned Data Scientist to join our Research Staff and help architect an industrial-scale data factory for next-generation speech and language AI foundation models. Conversational audio presents complex engineering and infrastructure constraints that are orders of magnitude harder than working with text due to high dimensionality, acoustic diversity, and scarcity of labeled data. In this cross-functional research capacity, you will explore new sequence modeling paradigms, design high-throughput preparation pipelines from a blank page, and build advanced benchmarking systems to capture nuanced context within human speech.

Key Responsibilities

• Data Factory Pipeline Architecture: Drive high-performance data acquisition, preparation, characterization, and transformation pipelines to supply training nodes for advanced speech models.
• Signal Processing Characterization: Develop complex characterizations of multi-channel conversational audio utilizing a diverse toolkit of deep learning architectures and statistical methods.
• DataOps Human-in-the-Loop Scaling: Collaborate with infrastructure teams to build automated systems that scale the capacity of human annotators to label high-value training assets.
• Benchmarking Methodology Design: Construct advanced verification frameworks, validation datasets, and evaluation metrics for tracking conversational voice platform performance.
• AI-First Workflow Automation: Actively integrate and build cutting-edge artificial intelligence tools to accelerate internal data experiments and amplify personal execution velocity.

Required Skills & Qualifications

• Demonstrated professional track record as a Data Scientist or Machine Learning Researcher building and owning complex data processing pipelines from the ground up.
• Deep theoretical and practical expertise applying statistical modeling and deep learning architectures to uncomplicate messy, high-dimensional datasets.
• Strong, production-grade software engineering skills in Python paired with extensive model training experience in PyTorch .
• Excellent communication skills with an innate ability to translate abstract scientific concepts into clear terms for multi-disciplinary product and engineering audiences.
• Location Context: 100% remote-first workspace flexibility open to qualified research staff permanently residing anywhere within the United States .

Preferred Strategic Indicators (Nice to Have)

• Advanced academic background or research specialization within Physics, Mechanical Engineering, or Natural Language Processing (NLP) .
• Direct hands-on experience building, training, or fine-tuning models specifically within the speech, acoustics, and audio AI domains.

What We Offer

• Estimated Base Salary Range: $150,000 – $220,000 USD per annum (Calibrated based on location coordinates and skills, supplemented by a 10% annual cash bonus and equity stock options).
• Comprehensive healthcare coverage including premium medical, dental, and vision insurance profiles.
• Unlimited Paid Time Off (PTO) structures paired with 12 paid US company holidays and flexible schedule routines.
• Generous paid parental leave, life insurance coverage, and short-term/long-term disability plans.
• Corporate 401(k) plan featuring an active company matching framework.
• Quarterly personal productivity stipends coupled with an educational budget and home office setup allowances.

Job Tags

Similar Jobs