AI Research Engineer

Overview

Research Engineer to characterize and measurably improve Emergent’s autonomous coding agents by building evaluation datasets, metrics, and production-grounded training experiments.

What you'll do

• Architect the next version of the agent by making core design choices.
• Characterize agent behavior and convert success/failure into quantitative metrics.
• Design and ship evaluations across reasoning, planning, tool use, code correctness, long-horizon execution, security, and reliability.
• Build evaluation datasets and methodologies, then validate against known signals and deploy dashboards.
• Run training and post-training experiments including fine-tuning, RLHF/RLAIF, DPO, distillation, reward modeling, prompt optimization, and judge-model calibration.
• Mine production data for failure modes, drive step-function capability gains, and own outcomes end-to-end.

What you'll need

• 5–8 years of AI experience, specifically training/fine-tuning models or designing rigorous evaluation/measurement systems.
• Hands-on modern AI stack experience and fluency in Python for research workflows.
• Experience with transformers and agent fine-tuning approaches such as RLHF/DPO/RL.
• Comfort with evaluation frameworks (e.g., Inspect, lm-eval-harness, or equivalent) and prompt optimization.
• Ability to defend benchmark and metric validity and assess statistical significance.
• Comfort reasoning about noise floors, confounds, distribution shift, judge bias, and selection effects.
• Independent operator skills with leadership presence and fast, rigorous execution.