Structured catalog
Evidence-based therapeutic models are held as structured, queryable data rather than buried in documents.
← All systemsChild Welfare
Service Matching Engine
A matching engine that scores child-welfare cases against a structured catalog of evidence-based therapeutic models, recommends the best fit, and shows the reasoning behind it.
In short
The Service Matching Engine scores child-welfare cases against a structured catalog of evidence-based therapeutic models and recommends the best-fit model with its reasoning shown. The catalog is held as structured, queryable data rather than buried in documents, so every case is scored against the same criteria. Each recommendation comes with the reasoning behind it, so it supports the caseworker instead of hiding the logic, and the final decision stays human-led. The result is consistent, explainable matching across cases that carry real consequences, applied the same way every time with the reasoning visible.
GPT-4oCustom MetadataLWCSalesforce
Best fitRecommended per case
ShownReasoning exposed
Human-ledFinal decision
The problem
Matching a family to the right evidence-based therapeutic model was manual and judgment-heavy, and the decisions carry real consequences.
Consistency across cases was hard to hold when every match was made from scratch.
How it works
Score the case, show the reasoning, keep the decision human-led.
route_to_human.pipeline Confident · auto Unsure · to a human
Incomingcases · docs
Modelconfidence
Human review
Flagged cases
Auto-resolved
Confident, cited
Score the case
Each case is scored against the catalog to recommend the best-fit model.
Show the reasoning
The recommendation comes with its reasoning, so it supports the caseworker instead of hiding the logic. The decision stays human-led.
What shipped
Consistent, explainable recommendations that support the decision instead of replacing the human making it.
The same catalog and logic are applied across cases, with the reasoning visible every time.
Book a call →