Machine Learning Engineer

Vindicor uses rules, expected-value logic, and machine learning to help chargeback teams decide which cases deserve attention. As a Machine Learning Engineer, you will improve those decision signals without turning the model into a black box.

The role focuses on evaluation, calibration, data quality, and production behavior. The output must be useful to an operator who needs to understand why a case was prioritized.

• Improve fight-or-pass scoring, calibration, and expected-value decision support
• Build reproducible training and evaluation datasets from dispute and outcome records
• Define offline metrics that reflect operational value, not only model accuracy
• Measure uncertainty, class imbalance, drift, and failure modes
• Expose model inputs and explanations in forms an operator can review
• Partner with backend and product engineering to ship and monitor model changes safely

• Improve fight-or-pass scoring, calibration, and expected-value decision support
• Build reproducible training and evaluation datasets from dispute and outcome records
• Define offline metrics that reflect operational value, not only model accuracy
• Measure uncertainty, class imbalance, drift, and failure modes
• Expose model inputs and explanations in forms an operator can review
• Partner with backend and product engineering to ship and monitor model changes safely

The model supports an operator decision; it does not hide or replace it. You will have direct access to the product roadmap and will be expected to document assumptions, failure modes, and evaluation evidence.

Compensation, equity, schedule, time off, equipment, and other employment terms are discussed directly during the interview process.