HW AI build out · AI For R*tards

HW AI build out

Built: AI Risk Stratification Tool for Homebirth Midwives

For Day 3, I built an AI-assisted risk screening and tracking tool designed for out-of-hospital midwifery care in Chat GPT & Claude

The tool:

screens maternal, obstetric, and current pregnancy factors
assigns weighted risk scores
classifies clients as: 🟢 Low risk 🟡 Moderate risk (review needed) 🔴 High risk (consult / transfer likely)
flags automatic red conditions
generates a clinical summary with next steps and chart-ready documentation

Why this tool:

Most midwives are tracking risk mentally, which creates inconsistency and makes it harder to clearly identify when someone is moving out of low-risk criteria.

Test Scenario:

Second-time mom, 32 weeks

Previous cesarean (wants VBAC)

Singleton pregnancy

No major medical conditions

Desires low-intervention birth

Feels only moderately supported by current provider

Result from the tool:

🟡 Moderate Risk — Needs Midwife Review

Key factors identified:

Prior cesarean (VBAC candidate)
Moderate provider alignment concerns

Clinical considerations:

Client may still be appropriate for out-of-hospital care depending on provider experience and support structure, but requires individualized assessment and planning.

Next step:

Recommend consultation to review VBAC support, provider regulations, and birth planning options.

What I learned:

The tool works best when applied to real scenarios I already understand. It creates clarity quickly and turns what is usually a mental process into something structured and repeatable grabbing data points that are normally in different places through out a clients medical chart. This can be used in real time during prenatal visits, inputting labs and ultrasounds, as well as for use in the labor records during a birth.

3 comments