Top AI & Machine Learning Jobs in San Francisco, CA

Leads Deepgram’s end-to-end Text-to-Speech research program, owning research strategy, technical direction, model development, evaluation, and production deployment. Advances neural audio modeling, prosody, expressiveness, multilingual generation, voice consistency, controllability, and inference performance. Builds and develops research teams, manages technical leaders, prioritizes experiments and compute, evaluates model quality, and partners with engineering and product leadership to ship production-grade models.

Own Deepgram’s end-to-end text-to-speech research program, including technical strategy, model architectures, experiments, training, evaluation, and production deployment. Lead researchers and technical managers, establish roadmaps, advance naturalness, expressiveness, multilingual generation, controllability, voice consistency, and inference efficiency, and partner with engineering and product leadership to ship high-quality models.

Research Staff will develop foundational voice AI technologies, including low-bitrate neural audio codecs, steerable speech generation, disentangled audio representations, latent recombination, synthetic audio data generation, and multimodal speech-to-speech models. The role also involves designing scalable architectures, training methods, and inference algorithms optimized for hardware, billion-hour datasets, and real-time deployment. Candidates need strong mathematical foundations, foundation-model expertise, large-scale data pipeline experience, rigorous experimentation skills, deployment optimization knowledge, and publications or open-source contributions in speech or language AI.

The Research Engineer will collaborate with scientists to develop and validate machine learning models for speech technologies, manage scalable training systems, and design tools for accessibility to non-technical users.

The role involves researching and developing large language models (LLMs) with a focus on transformer architecture, data curation, distributed training, and optimization. Responsibilities include conducting experiments, collaborating with teams, and staying updated on deep learning advancements.

As a Data Scientist at Deepgram, you'll tackle complex audio data challenges, develop scalable data pipelines, and collaborate with teams to advance voice AI technologies.

Build AI-powered workflows, agents, integrations, and internal tools that automate People Operations. Partner with People, Engineering, and IT to modernize onboarding and other employee processes, while implementing secure access controls, human oversight, escalation paths, and production-ready architectures. Prototype with users, iterate quickly, and measure improvements in efficiency, employee experience, and operational consistency.

Own and grow the self-serve developer funnel from first touch to paid customer. Define and ship product changes across onboarding, console, and activation. Build instrumentation, experimentation, and business rhythms; run rapid hypotheses, measure outcomes, and collaborate with engineering, design, dev rel, marketing, sales, and analytics.

Lead and build Deepgram's AI enablement function: prototype and ship reusable agents, MCP servers, paved-road workflows, prompt and retrieval patterns, and an enablement hub. Drive company-wide AI adoption with measurable productivity metrics, embed safe-use guardrails, partner across Platform, Security, Data, and People Ops, and grow a central team and distributed champions network.

Lead end-to-end delivery of ML infrastructure and AI tooling programs, defining architecture and rollout strategies for training, serving, and monitoring. Coordinate research, engineering, product, and data teams to optimize inference cost/latency, build internal tools for experiment tracking and model versioning, and resolve bottlenecks to enable faster, scalable ML deployments.