Member of Technical Staff, Machine LearningCNTR

CANDIDATES WITHOUT ACTIVE WORKING LINKEDIN PROFILE WILL NOT BE CONSIDERED! PLEASE ADD IT TO YOUR CV OR APPLICATION. THANK YOU.

Were looking for a Machine Learning Engineer to own the ML lifecycle end to end from understanding customer problems through designing datasets improving models building evaluation systems and shipping production pipelines that deliver measurable improvements in data quality. The work runs directly with frontier AI labs to understand difficult ambiguous data problems and then build systems that solve them at scale. One week might mean fine-tuning a multimodal model to improve recall on a hard edge case; the next might mean engineering a VLM-based QA pipeline designing a new evaluation framework or running a large-scale filtering pipeline across millions of hours of multimodal data. Youll own model quality for customer-facing video understanding problems build automated evaluation and QA pipelines on top of frontier models and design high-precision filtering ranking retrieval and labeling systems over internet-scale video datasets. This is a good fit for someone who enjoys owning problems end to end and wants their work to ship directly into the models defining the frontier.

Details:
Schedule: Full time
Location: Onsite (San Francisco CA 5 days/week)
Start: ASAP
Duration: Long-term

About the project:
The project is an AI research lab building some of the highest-quality multimodal datasets in the industry spanning video audio images text and 3D. It combines exabyte-scale data infrastructure novel multimodal understanding techniques and dozens of proprietary data sources to produce datasets that push the frontier of foundation models. Video makes up the majority of internet traffic and across modalities data has become the enabling medium behind creativity communication gaming AR/VR and robotics so the team is focused on solving the biggest bottleneck to growth in those applications: high-quality training data. The organization partners directly with the worlds top AI labs and operates at meaningful scale with a lean senior team. It recently raised a Series A from tier-1 venture firms. Its an environment where a small team ships work that lands directly in production models used by major AI labs.

You have:
Strong Python engineering skills with experience building production ML systems
Experience training fine-tuning or deploying modern deep learning models with hands-on PyTorch and modern foundation model work
Excellent intuition for evaluation dataset quality precision/recall tradeoffs and edge cases
A track record of owning projects end to end from customer problem through internal pipelines to deployed solution
Strong communication skills and comfort working directly with customers and cross-functional teams
Willingness to work onsite in San Francisco 5 days per week
Nice to have: direct experience with video multimodal AI or frontier foundation models
Nice to have: a track record of rapidly prototyping with new AI models and APIs
Nice to have: experience running large-scale data filtering or curation pipelines
Nice to have: background at an AI lab a frontier data company or a high-scale ML infrastructure team

What to do:
Own model quality for customer-facing video understanding problems
Fine-tune vision-language and multimodal foundation models for specialized tasks
Build automated evaluation and QA pipelines on top of frontier models (Gemini GPT Claude) and open-source VLMs
Design high-precision filtering ranking retrieval and labeling systems over internet-scale video datasets
Build and maintain the datasets benchmarks and evaluation frameworks that continuously improve model quality
Build production ML pipelines spanning preprocessing inference post-processing and quality validation
Work directly with frontier AI labs to translate ambiguous requirements into scalable systems