What are the responsibilities and job description for the Applied Scientist position at Stealth Talent Solutions?
About the Company
We're a fast-growing adtech company building large-scale predictive systems that power real-time optimization across mobile and connected TV advertising. Our machine learning team in New York is expanding, and we're looking for sharp, research-minded early-career scientists to grow with us.
About the Role
As a Junior Applied Scientist, you'll sit at the intersection of research and production engineering — translating statistical and ML theory into scalable systems that make measurable impact from day one. You'll work alongside senior applied scientists and engineers on problems involving high-velocity data, real-time inference, and probabilistic modeling at scale.
This is a role for someone who thinks like a researcher but wants to ship things.
What You'll Do
• Build and improve predictive models that operate in real-time production environments at massive scale
• Mine large-scale datasets to surface patterns, signals, and features that improve model performance
• Design, run, and interpret experiments (A/B tests, holdouts, causal inference) to validate hypotheses
• Write clean, well-tested, production-grade Python code integrated into ML pipelines
• Collaborate with senior scientists and engineers to tackle complex modeling challenges
• Monitor and analyze live model performance, identifying degradation and improvement opportunities
What We're Looking For
• MS or PhD in Computer Science, Statistics, Applied Mathematics, Operations Research, or a related quantitative field — or equivalent research/industry experience
• 0–3 years of experience; new grads, PhD candidates, and researchers coming out of top-tier internships (e.g., Google, Meta, Amazon, Microsoft Research) are strongly encouraged to apply
• Strong foundations in probability, statistics, and machine learning (not just familiarity — you can derive it)
• Proficiency in Python; experience with ML frameworks (PyTorch, TensorFlow, JAX, scikit-learn)
• Experience working with large datasets and an intuition for what makes a good feature
• A research mindset: you ask "why does this work" before "does this work"
• Comfort operating in ambiguous, fast-moving environments
Nice to Have
• Experience with real-time or online learning systems
• Familiarity with probabilistic modeling, Bayesian methods, or causal inference
• Published research or strong internship track record at a research-driven org
We are an equal opportunity employer and value diversity at our company.