Research Engineer improves world models by curating multimodal (image/video) training datasets at scale. Builds processing pipelines, deploys ML for enrichment, and closes data-model-evaluation loop using strong engineering and research skills.
Salary not listed
On-siteML Engineering
About the role
What You’ll Do
Discover, evaluate, and acquire training data. Write scrapers, work with APIs, and make judgement calls about sources.
Build data processing and curation systems. Design pipelines for filtering, deduplication, quality scoring, and curation.
Look at the actual data constantly. Sample outputs, spot distributional issues like too many screenshots or low-resolution crops.
Close the data → model → evaluation loop. Diagnose model failures, trace to data issues, and design fixes.
Deploy ML models for data enrichment (captioning, quality scoring, text embedding, segmentation, classification). Evaluate their impact.
Make systematic, documented decisions. Ensure reproducibility for thresholds, criteria, and ratios.
What We Require
Strong software engineering fundamentals. Well-abstracted, readable code and reusable tools.
Deep experience with image and video data at scale. Data formats, processing libraries (OpenCV, PIL, FFmpeg, PyAV).
Experience with distributed computing. Frameworks like Apache Beam, Spark, Kubernetes, Ray.
Experience using ML models as components. Build inference pipelines at billion scale and evaluate outcomes.
Research-oriented approach to data decisions. Design experiments to validate choices.
Familiarity with the model training lifecycle. Understand data composition effects.
Obsession for the data-model-evaluation loop with proven track record.
What We’d Love To See
Familiarity with columnar/large-scale data storage (PyArrow, Lance, Vortex, DeepMind Bagz).
Track record of discovering/integrating new data sources.
Direct experience closing data → model quality loop.
Strong visual intuition for data quality/diversity.
Characterize, analyze, and optimize performance of state-of-the-art AI models on Cerebras' wafer-scale hardware. Build performance models, optimize kernels and compilers, debug runtime behavior, and develop visualization tools to influence next-gen AI architecture.
Salary not listed
On-site3+ YOEML Engineering
Research Engineer, Privacy
OpenAISan Francisco, CA
Research Engineer on OpenAI's Privacy team designing and prototyping privacy-preserving ML algorithms like differential privacy and federated learning at scale. Requires hands-on PETs experience, fluency in PyTorch/JAX, and a track record implementing or publishing novel privacy work.
380k – 445k/yr
HybridML Engineering
Research Engineer
ConsoleSan Francisco, CA
Research Engineer building self-improving AI agent systems at Console. Develop eval/optimization loops, fine-tune specialist models, and improve agent reasoning over enterprise context using production data to drive measurable gains in quality, latency, and reliability.
200k – 350k/yr
On-siteML Engineering
Software Engineer, AI Platform
NotionSan Francisco, CA +1
Build and scale the shared AI platform foundations at Notion, enabling fast and safe shipping of AI products. Requires experience with LLM/ML platforms, strong ownership, and comfort across backend, infrastructure, and product code.
180k – 201k/yr
Hybrid5+ YOEML Engineering
Machine Learning Engineer
LiftoffCalifornia
Machine Learning Engineer building statistical models, optimization systems, and experiments for mobile ad tech economics on the Revenue Engine team. Requires PhD in CS/ML/Economics and industry experience applying ML or economics at scale.