Sieve
Member of Technical Staff, Machine Learning
San Francisco · Staff+
Sponsorship not specifiedDetected 9 days ago
PythonSwiftMachine LearningDeep LearningPyTorchNLPMLOpsAR/VRResearchCommunication
About the role
- One week you might fine-tune a multimodal model to improve recall on a difficult edge case.
Responsibilities
- Own model quality for customer-facing video understanding problems
- Build automated evaluation and QA pipelines using frontier models like Gemini, GPT, Claude, and open-source VLMs
- Design high-precision filtering, ranking, retrieval, and labeling systems over internet-scale video datasets
- Create datasets, benchmarks, and evaluation frameworks that continuously improve model quality
- Develop production ML pipelines spanning preprocessing, inference, post-processing, and quality validation
- Strong Python engineer with experience building production ML systems
Benefits
- Fine-tune vision-language and multimodal foundation models for specialized tasks
- Experience training, fine-tuning, or deploying modern deep learning models
Company info
- Sieve is an AI research lab building the world's highest-quality multimodal datasets - spanning video, audio, images, text, and 3D. We combine exabyte-scale data infrastructure, novel multimodal understanding techniques, and dozens of proprietary data sources to develop datasets that push the frontier of foundation models. Video alone makes up 80% of internet traffic, and across modalities, data has become the enabling medium powering creativity, communication, gaming, AR/VR, and robotics. Sieve exists to solve the biggest bottleneck in the growth of these applications: high-quality training data.
- We've partnered with the world's top AI labs and did $XXM last quarter alone, as a team of just ~25 people. We also raised our Series A from Tier 1 firms such as Matrix Partners, Swift Ventures, Y Combinator, and AI Grant.
- Sieve is one of the most capital-efficient teams in AI - roughly 25 people serving the world's leading AI labs across every major data modality. You'll join early, own problems end-to-end, and watch your work ship directly into the models defining the frontier.
- Sieve is an AI research lab building the world's highest-quality multimodal datasets - spanning video, audio, images, text, and 3D.
- We combine exabyte-scale data infrastructure, novel multimodal understanding techniques, and dozens of proprietary data sources to develop datasets that push the frontier of foundation models.
- Video alone makes up 80% of internet traffic, and across modalities, data has become the enabling medium powering creativity, communication, gaming, AR/VR, and robotics.
- Sieve exists to solve the biggest bottleneck in the growth of these applications: high-quality training data.
- We've partnered with the world's top AI labs and did $XXM last quarter alone, as a team of just ~25 people.
- We also raised our Series A from Tier 1 firms such as Matrix Partners, Swift Ventures, Y Combinator, and AI Grant.
- Sieve is one of the most capital-efficient teams in AI - roughly 25 people serving the world's leading AI labs across every major data modality.
- You'll join early, own problems end-to-end, and watch your work ship directly into the models defining the frontier.
This listing is sourced directly from Sieve's careers page and normalized into a canonical job model.