About the Company (Confidential)
Our client is a cutting-edge AI research company specializing exclusively in video data —one of the fastest-growing and most technically demanding domains in machine learning. With just 12 team members , the company has already :
- Partnered with leading AI labs ,
- Achieved multi-million dollar revenue last quarter , and
- Recently raised a Series A from top-tier investors , including Matrix Partners, Swift Ventures, Y Combinator, and AI Grant .
They are solving one of the biggest bottlenecks in AI today : creating high-quality, scalable training datasets for video modeling . Their work spans exabyte-scale infrastructure, novel video understanding techniques, and multi-modal data pipelines across video, audio, and text.
This is a rare opportunity to join an elite, deeply technical, and extremely high-leverage team at the frontier of AI research.
The Opportunity
As an Applied Research Engineer , you will build the high-performance infrastructure, pipelines, and research-driven components that push forward the state of video understanding at internet scale. This is an in-person role at the San Francisco HQ, working closely with founders and senior researchers.
You’ll operate end-to-end across ambiguous, open-ended research problems—designing clever pre / post-processing systems, optimizing inference, parallelizing large-scale workloads, and contributing to the core technical foundation of next-generation video modeling.
This is a high-ownership role ideal for someone who thrives in low-structure environments, loves working directly with models and APIs, and enjoys extracting every ounce of performance through deep problem-solving.
What You’ll Do
Build and optimize video, audio, and text processing pipelines at massive scale.Work directly with models, APIs, and custom systems to push performance boundaries.Develop high-precision video understanding building blocks for research and production.Introduce techniques for parallelism, pipelining, inference optimization , and clever pre / post-processing.Collaborate with customers and external teams to translate real-world needs into scalable solutions.Contribute to the advancement of video modeling and dataset creation for top AI labs.Operate across ambiguous research spaces—rapidly designing experiments, prototypes, and iterations.Requirements
2+ years of experience in computer vision or audio processing .Strong Python engineering experience; hands-on with PyTorch or similar ML frameworks.Excellent communication skills, especially with customers or external-facing partners.Writes clean, maintainable, production-quality code.Comfortable working with multi-modal models and APIs.Deep interest in video, media technologies, and high-performance system design .Motivated by building full end-to-end systems—not just model training.Strong ability to break down problems from customer impact → technical building blocks.In-person at the San Francisco HQ.Bonus Points :
Active open-source contributor.Experience as an early hire at a fast-moving startup.GitHub portfolio or public technical projects.Compensation & Benefits
Salary : $200,000–$250,000Equity : 0.05%–0.4%Employment : Full-time, in-person SFHealth insuranceFree lunch & dinnerCompetitive total compensation package