SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Cinder is mission-critical infrastructure for content moderation and abuse prevention at scale. The company helps the world's largest internet platforms fight fraud, abuse, and manipulation through AI-powered policy enforcement, real-time agent deployment, threat investigation, and safety program validation. Backed by Accel and Y Combinator, Cinder is a small, fast-moving team solving real customer problems in content safety.
You'll join as an AI/ML Engineer to build production ML systems that power Cinder's core platform. Your responsibilities span the full ML lifecycle: transforming messy customer data into trainable datasets, selecting the right model approach (classical classifiers, fine-tuned LLMs, or third-party APIs based on cost/latency/accuracy tradeoffs), and shipping models that directly impact customer outcomes.
Key focus areas include improving classification pipelines and detection strategies for harmful content, developing intelligent features for moderators, building evaluation and metrics infrastructure, and partnering with the founding data scientist and data engineer to shape agent evaluation architecture and data infrastructure. You'll also mentor teammates as Cinder's ML capability matures.
The ideal candidate has 5–8+ years of ML engineering experience on small teams, with a proven track record shipping production systems. You've taken classification problems from raw, unlabeled data to production, understand when LLMs justify their cost versus classical models, and have built ML infrastructure from scratch (training pipelines, serving, evaluation, monitoring). You're fluent in fundamentals: feature engineering, train/test methodology, metric selection for imbalanced data, and principled hyperparameter tuning. Strong Python skills with PyTorch, scikit-learn, XGBoost, and LangChain are essential. You have solid MLOps foundations (CI/CD, versioning, experiment tracking, drift detection, monitoring) and experience designing inference systems with explicit latency/throughput targets. AWS and Terraform experience is a plus.
Cinder is based in NYC and will relocate for this role, emphasizing in-person collaboration with at least two all-company events annually. Benefits include health/vision/dental, 401(k) with matching, commuter benefits, and a fully stocked office with paid meals.
