Deeter Analytics
Machine Learning Engineer
About this role
Join an investment research firm as a Machine Learning Engineer to build deep-learning models on market data end-to-end, from raw data through daily production systems. You'll own the complete loop—data, models, compute, and evaluation—working remotely with a senior researcher to turn desk ideas into actionable trading insights.
What you'll do
- Build and train deep-learning models on market data, from baseline to production
- Manage GPU compute infrastructure (local or AWS) including environment setup, containers, and cost optimization
- Design leakage-proof validation and regime-aware testing on time-ordered financial data
- Read and reproduce recent research in foundation models, time-series, and reinforcement learning
- Own the entire project lifecycle from raw data ingestion to models that inform daily trading decisions
- Use modern AI tools for code generation, literature review, and data wrangling while verifying results
What they're looking for
- Deep learning (optimization, initialization, normalization, attention mechanisms)
- Linear algebra, probability, and statistics
- Python and PyTorch or TensorFlow
- GPU compute and cloud infrastructure (AWS)
- Time-series modeling and financial data analysis
- Experiment tracking and rigorous validation methodology
- Docker and containerization
- Self-directed learning and rapid prototyping
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Deeter Analytics
Deeter Analytics builds foundational AI models for financial applications using multimodal datasets. The company is hiring research engineers and machine learning specialists who can drive work from research and prototyping through production deployment.
View all jobs at Deeter AnalyticsLikely interview questions
- Walk us through a personal machine learning project you built—what was the hardest part and how did you solve it?
- Describe a time when your model's results looked too good. How would you investigate whether it was real or noise?