Applied ML Engineer for Educational DS System
Budget: $50 – $0 USD
We are building a high-stakes educational decision-support system, not a generic AI product.
This role is for an Applied ML Engineer who is comfortable working with uncertainty, calibration, and small/noisy data, and who understands that accuracy alone is not enough.
You will work closely with learning scientists and a psychometrician to implement and validate Bayesian learner models that support a single, bounded decision: identifying prerequisite readiness gaps while explicitly representing uncertainty.
If your instinct is to “just add a bigger model,” this role is not for you.
What You Will Do
- Implement and maintain Bayesian learner models (e.g., BKT, IRT, related probabilistic approaches)
- Explicitly manage uncertainty and calibration (false positives / false negatives matter)
- Integrate models with assessment logic and decision rules
- Support model validation and evaluation (not just training)
- Work within strict architectural constraints defined by a systems lead
- Collaborate closely with a psychometrician and education experts
Required Background
- MSc or PhD in Computer Science, Statistics, Applied Mathematics, or related field
- Strong grounding in probability and statistical modeling
Required Experience
- 5+ years applied machine learning experience
- Demonstrated experience with probabilistic or Bayesian models
- Experience with model calibration, error analysis, and evaluation
- Comfort working with limited, imperfect, or human-generated data
- Explicitly NOT What We’re Looking For (Read Carefully)
This role is for an Applied ML Engineer who is comfortable working with uncertainty, calibration, and small/noisy data, and who understands that accuracy alone is not enough.
You will work closely with learning scientists and a psychometrician to implement and validate Bayesian learner models that support a single, bounded decision: identifying prerequisite readiness gaps while explicitly representing uncertainty.
If your instinct is to “just add a bigger model,” this role is not for you.
What You Will Do
- Implement and maintain Bayesian learner models (e.g., BKT, IRT, related probabilistic approaches)
- Explicitly manage uncertainty and calibration (false positives / false negatives matter)
- Integrate models with assessment logic and decision rules
- Support model validation and evaluation (not just training)
- Work within strict architectural constraints defined by a systems lead
- Collaborate closely with a psychometrician and education experts
Required Background
- MSc or PhD in Computer Science, Statistics, Applied Mathematics, or related field
- Strong grounding in probability and statistical modeling
Required Experience
- 5+ years applied machine learning experience
- Demonstrated experience with probabilistic or Bayesian models
- Experience with model calibration, error analysis, and evaluation
- Comfort working with limited, imperfect, or human-generated data
- Explicitly NOT What We’re Looking For (Read Carefully)