Foundational ML Research Lead – Finance & RL

Enigma; New York (NY); Hybrid; USD 180,000 - 240,000; Full time

Job summary

Enigma is hiring a Foundation Model Training Lead to own the core models, overseeing end-to-end development from pretraining through optimization in a deeply technical, high-impact role connected to financial systems.

Based in New York with 3–4 days in-office, you will lead end-to-end training of foundation models, design MoE and data strategies, and collaborate with a small, fast-moving team to push research toward production-ready systems.

Qualifications

• Proven experience training large-scale models end-to-end (not just fine-tuning).
• Strong background in deep learning and large model architectures.
• Experience with reinforcement learning in real-world or production settings.
• Hands-on work with MoE architectures and/or distributed training systems.
• Deep understanding of training dynamics, scaling laws, and optimization; data quality and curation for large models.

Responsibilities

• Lead the end-to-end training of large-scale foundation models.
• Design and implement pretraining and continued training strategies on financial data.
• MoE design and routing implementation.
• Tokenization strategies for financial data.
• Build RL training loops tied to real-world trading performance (P&L).
• Develop systems for training stability, scaling, and performance optimization.
• Define and execute data strategy (dataset construction, curation, filtering, labeling).
• Work closely with engineering to build scalable training infrastructure.
• Contribute to broader research direction and technical roadmap.

Skills

Large-scale models

Deep learning

Reinforcement learning

MoE architectures

Distributed training

Data quality & curation

Job description

Enigma is hiring a Foundation Model Training Lead to own the core models, overseeing end-to-end development from pretraining through optimization in a deeply technical, high-impact role connected to financial systems.

Based in New York with 3–4 days in-office, you will lead end-to-end training of foundation models, design MoE and data strategies, and collaborate with a small, fast-moving team to push research toward production-ready systems.