Deepgram is the leading platform underpinning the emerging trillion-dollar Voice AI economy, providing real-time APIs for speech-to-text (STT), text-to-speech (TTS), and building production-grade voice agents at scale. More than 200,000 developers and 1,300+ organizations build voice offerings that are ‘Powered by Deepgram’, including Twilio, Cloudflare, Sierra, Decagon, Vapi, Daily, Cresta, Granola, and Jack in the Box. Deepgram’s voice-native foundation models are accessed through cloud APIs or as self-hosted and on-premises software, with unmatched accuracy, low latency, and cost efficiency. Backed by a recent Series C led by leading global investors and strategic partners, Deepgram has processed over 50,000 years of audio and transcribed more than 1 trillion words. There is no organization in the world that understands voice better than Deepgram.
At Deepgram, we expect an AI-first mindset—AI use and comfort aren’t optional, they’re core to how we operate, innovate, and measure performance.
Every team member who works at Deepgram is expected to actively use and experiment with advanced AI tools, and even build your own into your everyday work. We measure how effectively AI is applied to deliver results, and consistent, creative use of the latest AI capabilities is key to success here. Candidates should be comfortable adopting new models and modes quickly, integrating AI into their workflows, and continuously pushing the boundaries of what these technologies can do.
Additionally, we move at the pace of AI. Change is rapid, and you can expect your day-to-day work to evolve just as quickly. This may not be the right role if you’re not excited to experiment, adapt, think on your feet, and learn constantly, or if you’re seeking something highly prescriptive with a traditional 9-to-5.
At Deepgram, data isn’t just fuel for our models, it’s a product in its own right. Our vertically integrated voice AI platform depends on strategically created, curated, and labeled audio: from original data collection and augmentation, to structured workflows and evaluation sets. These data pipelines support a range of cutting-edge technologies, including audio intelligence models, conversational voice agents, STT, and TTS.
We’re looking for a hands-on, systems-minded Program Manager to lead the design and execution of various voice data programs. This role is ideal for someone who thrives on building from scratch, someone who can take an abstract modeling goal or product need and turn it into a concrete data strategy and pipeline with tools, guidelines, and quality safeguards in place. This requires ownership to understand frontier research strategies, building custom style guides, prototyping new tools, and directly influencing how data shapes our products.
This is a role for builders, someone who can spot a gap, roll up their sleeves, and design the solution. You’ll be at the center of Deepgram’s model development cycle, working across Research, Engineering, and Product, and you’ll be elbow-deep in both the day-to-day execution and the systems thinking required to scale it.
This role reports to the VP of Data Operations.
Design, launch, and own end-to-end data workflows: from raw audio ingestion to production-ready datasets
Build and evolve labeling specs, style guides, and instructional documentation for global annotation teams
Identify opportunities for better tooling, automation, and workflow optimization, and lead their implementation
Translate product goals and model requirements into data creation strategies, deciding what to build, how to build it, and why it matters for product impact
Own the full lifecycle for your domains — customer expectations, data acquisition, preparation, scaling, provenance, and evaluation — and be accountable for the model outcome, not just the data hand-off.
Prototype and deploy data tools and infrastructure
Collaborate with Research and Engineering to align data collection with model training architecture and downstream product impact
Track advancements in speech AI research and evolving market use cases to inform labeling approaches and data design priorities
Partner with QA and Evaluation leads to deliver high-quality, human-in-the-loop datasets and benchmarks
Manage and mentor data vendors, freelancers, and potentially internal ICs as the team grows
Track throughput, data quality, and vendor performance
Drive continuous improvement in speed, cost-efficiency, and quality across all data operations
Curate and refine datasets to align with specific product goals, linguistic coverage, or research hypotheses
Experience owning data, ML, or operations programs end-to-end in program/project or product management.
Fluency working directly with technical teams; you can hold a conversation about data quality, evaluation, and model impact.
Systems thinking, understanding how decisions propagate across a system, reason up from fundamentals instead of defaulting to convention, and design solutions that hold up as things scale.
A track record of prioritization under constraint — deciding what to fund, what to cut, and how to sequence it.
Strong operating instincts: you scope, sequence, assign, and ship, and nothing stalls because someone didn't know the next step.
Demonstrated ability to design and build scalable processes, not just manage existing ones
Direct exposure to speech/audio, ASR, or TTS data, and the specific nuances of multilingual, code-switched, low-resource, or domain-specific data.
Experience with active-learning or data-selection approaches
Startup or high-ambiguity experience
Believe that data is a product, not just a resource, and you want to help define what great voice data looks like
Enjoy turning experimental ideas into robust, repeatable systems that can scale to production
Thrive in ambiguity and take initiative without waiting for perfect specs, preferring action over perfection and iteration over indecision
