SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Deepgram is the leading platform for Voice AI, providing real-time APIs for speech-to-text, text-to-speech, and voice agents at scale. The company serves 200,000+ developers and 1,300+ organizations including Twilio, Cloudflare, and Vapi, and has processed over 50,000 years of audio.
As a Software Test Engineer, you will design, build, and maintain comprehensive automated test frameworks and exploratory test suites across Deepgram's products, models, APIs, and data platforms. You'll translate product requirements and model metrics into automated regression tests, evaluation pipelines, data-quality gates, load tests, and release criteria.
Key responsibilities include:
- Define and execute well-designed test plans across products, APIs, SDKs, and model-powered features
- Design and maintain automated test suites for functional, integration, end-to-end, regression, API, and service-level testing
- Build representative, customer-focused, and adversarial test datasets and test environments
- Validate AI-powered behavior (speech-to-text, text-to-speech) using appropriate metrics and human review
- Build testing infrastructure including test harnesses, dashboards, and result-aggregation pipelines
- Integrate automated tests into CI/CD pipelines for continuous regression detection
- Partner with Engineering, Product, Research, and Data teams to understand system behavior and establish test coverage
- Test data ingestion, processing, and quality-control workflows
- Execute staging, production, load, and user acceptance testing
- Write precise, actionable bug reports with reproducible steps and clear severity
- Help raise engineering and QA standards through code reviews and technical discussions
You'll work in an AI-first environment where rapid experimentation, continuous learning, and creative AI tool adoption are core to success. The role requires comfort with evolving requirements and a mindset geared toward breaking systems and probing edge cases.
