3 months ago
Base Salary
$190k - $320k/yr
Responsibilities
- Design, build, and automate offline and live evaluation pipelines for speech and multimodal models.
- Create safe, versioned, privacy-aware tooling for dataset curation and discovery.
- Prototype, train, fine-tune, and deploy state-of-the-art voice models in partnership with research and infrastructure teams.
- Scale training and inference for LLM-class workloads while optimizing latency, throughput, and cost.
- Build monitoring and live evaluations to detect quality regressions before they affect users.
- Move research ideas from prototype to production rapidly with appropriate tests and dashboards.
- Set technical direction, drive consensus, and collaborate with product, infrastructure, and research teams.
Requirements
- Expert-level PyTorch experience.
- Production software engineering experience with the ability to write code across the stack.
- Hands-on experience training or fine-tuning large language models or other large-scale models using varied techniques.
- Expertise designing evaluation metrics and harnesses that predict user experience.
- Deep knowledge of the ML lifecycle, including dataset operations, training pipelines, evaluation frameworks, deployment, and monitoring.
- History of shipping complex, user-facing online ML systems to production despite changing requirements and obstacles.
- High agency and sound judgment when working independently or collaboratively.
- Track record of setting technical direction, driving consensus, and partnering effectively across product, infrastructure, and research.
Benefits
- Full-time employee position.
- 401(k) with a maximum employer match of 3.5% of compensation.
- 100% employer-paid health, vision, and dental benefits for employees and dependents.
- Unlimited paid time off and sick time.
- Flexible spending account with employer matching up to $1,650 per year for a medical FSA.
- Guardian Employee Assistance Program.
- Competitive stock options and an opportunity to share in the company's success.