ML Ops Infrastructure Engineer

Deepgram is a leading Voice AI platform providing real-time APIs for speech-to-text, text-to-speech, and production-grade voice agents. Over 200,000 developers and 1,300+ organizations build on Deepgram, including Twilio, Cloudflare, and Sierra. Deepgram's voice-native foundation models are available via cloud APIs or self-hosted/on-premises, with a focus on accuracy, low latency, and cost efficiency.

As an ML Ops Infrastructure Engineer, you will own the bridge between research and production, building pipelines, deployment systems, and testing infrastructure to take models from experimental to battle-tested at scale.

• Design and build CI/CD pipelines for ML model development, validation, and deployment.
• Architect and maintain model deployment pipelines from research to staging to production.
• Build A/B testing infrastructure for controlled rollouts and performance measurement.
• Implement monitoring for model performance: accuracy, latency, drift detection, and regression alerts.
• Develop automated retraining pipelines triggered by data changes or performance degradation.
• Create build and test environments that mirror production.
• Establish model versioning, artifact management, and rollback capabilities.
• Collaborate with research engineers to enforce model quality gates.
• Build observability dashboards for model health.
• Optimize model serving infrastructure for latency, throughput, and cost.

• 4+ years of experience in MLOps, DevOps, or infrastructure engineering with a focus on ML systems.
• Strong proficiency in Python and building automation for ML workflows.
• Deep experience with CI/CD systems and pipelines.
• Hands-on experience with Docker and Kubernetes.
• Practical experience deploying and serving ML models in production.
• Familiarity with model evaluation, validation, and QA processes.
• Understanding of monitoring and observability for ML systems.
• Strong problem-solving skills and bias toward automation.

• Experience with model serving frameworks like NVIDIA Triton, TensorRT, or ONNX Runtime.
• Background in speech, audio, or real-time media ML.
• Experience with Infrastructure as Code (Terraform, Pulumi).
• Hands-on experience with monitoring stacks (Prometheus, Grafana, Datadog).
• Familiarity with GPU-accelerated inference optimization.
• Experience with feature stores, data versioning, or ML metadata management.
• Knowledge of canary deployment strategies.

Salary range: $160,000 - $220,000 USD. Equity and bonus offered.

Apply via the provided application link.