Applied Research Engineer – Video Data & ML

About Turing

Based in Palo Alto, California, Turing is one of the world's fastest-growing AI companies accelerating the advancement and deployment of powerful AI systems. Turing helps customers in two ways: working with the world’s leading AI labs to advance frontier model capabilities in thinking, reasoning, coding, agentic behavior, multimodality, multilingualism, STEM and frontier knowledge; and leveraging that expertise to build real-world AI systems that solve mission-critical priorities for Fortune 500 companies and government institutions. Turing has received numerous awards, including Forbes's "One of America's Best Startup Employers," #1 on The Information's annual list of "Most Promising B2B Companies," and Fast Company's annual list of the "World's Most Innovative Companies." Turing's leadership team includes AI technologists from industry giants Meta, Google, Microsoft, Apple, Amazon, Twitter, McKinsey, Bain, Stanford, Caltech, and MIT. For more information on Turing, visit www.turing.com. For information on upcoming Turing AGI Icons events, visit go.turing.com/agi-icons.

Overview

We are seeking an Applied Research Engineer with a strong foundation in video generation, machine learning, or computer vision to help improve the quality of datasets and workflows powering next-generation video synthesis models. This role is ideal for a candidate with 3–5 years of experience in ML/AI who is eager to grow their expertise through hands-on data development and small-model fine-tuning under the mentorship of senior researchers.

You’ll work with ML teams, QA leads, and delivery managers to curate high-quality training data for generative models, contribute to targeted model experiments, and refine workflows for controllable and high-fidelity video generation. Strong cross-functional communication is essential to transform modeling requirements into actionable data specs and evaluation criteria.

Key Responsibilities

Data Curation for Video Generation

• Co-develop structured guidelines for datasets powering generative models, with an emphasis on:
• Prompt-conditioned video generation (text-to-video, image-to-video)
• Consistency in motion, temporal coherence, and visual quality
• Control signal labeling (e.g., camera trajectories, depth, optical flow)
• Style, identity, or scene constraints across frames
• Collaborate with ML stakeholders to align data specs with generation tasks such as video inpainting, editing, expansion, or simulation-to-real synthesis.

Benchmark-Driven Data Optimization

• Analyze where dataset limitations are causing poor outputs or failure cases in generative benchmarks (e.g., Video-Bench, VBench, or internal video synthesis metrics).
• Recommend guideline updates and synthetic data augmentations based on output inspection and metric evaluations (FID, IS, CLIP similarity, etc.).

Model Collaboration & Fine-Tuning

• Assist in fine-tuning or conditioning small video generation models (e.g., diffusion-based or transformer-based architectures) under the guidance of senior engineers.
• Run targeted experiments to evaluate model behavior across data variations and prompt types.

QA and Evaluation Process Support

• Build structured review protocols for evaluating generated video content, including:
• Artifact detection
• Temporal jitter or drift
• Semantic coherence with prompts
• Help define feedback loops for iterative data refinement and edge-case tracking.

Cross-Functional Communication

• Act as a liaison between ML engineers, data operations, and evaluation teams to ensure dataset alignment with generative objectives.
• Produce clear documentation and updates on data design, annotation strategy, and model performance feedback.

Qualifications

• 3–5 years of experience in computer vision, applied ML, or generative AI, especially with video, image, or temporal modeling.
• Familiarity with video synthesis techniques, including diffusion models, transformer-based generators, or GAN variants.
• 3+ years of experience writing production-quality software, preferably in machine learning, AI, or data science contexts.
• Proficiency in Python and familiarity with libraries such as PyTorch, Keras, scikit-learn, and Hugging Face.
• Hands-on experience with basic fine-tuning or evaluation of generative models (e.g., with PyTorch, TensorFlow, Hugging Face, or Runway).
• Exposure to tools or platforms for dataset curation and video inspection (e.g., CVAT, custom viewers, or synthetic data generators).
• Strong grasp of the data lifecycle in generative AI—annotation, prompt engineering, synthetic data usage, and evaluation.
• Ability to read ML/AI research papers (e.g., on video diffusion, text-to-video, or controllable generation) and apply insights to dataset or model design.
• Excellent communication skills—confident in presenting technical findings, coordinating with stakeholders, and translating between research and delivery teams.

What Success Looks Like

• Curated datasets that enhance fidelity, consistency, and controllability in generated video outputs.
• Documented guidelines and QA protocols that improve model outputs across real-world tasks and benchmarks.
• Targeted fine-tuning or conditioning efforts that lead to measurable improvements in generation quality.
• Seamless collaboration across ML, data, and QA teams, accelerating the development and deployment of video generation systems.

Compensation: $170,000 to $200,000 + Equity

Advantages of joining Turing: