Empowering Diversion with Data Science, Machine Learning -A Juvenile Court Case Study

Challenges

The client had diversion data going back to 2016 but they were data rich and information poor.

• Manual outcome tracking and no effective way of linking diversion outcomes to interventions.
• No visibility into which interventions were working or not working.
• Lack of insights resulted in an inconsistency in decision making.
• Inability to link staff actions to outcomes and understand drivers of outcomes.

Solution

eMoksha's Data Science Service-based diversion solution was implemented.

• Enabled the court to set benchmarks/performance measures, measure metrics and outcomes, and assess the effectiveness of the interventions on the outcomes.
• Was expanded to include a diversion data entry platform to have rules-based diversion data entry.

Impact

The diversion program was completely transformed. A number of program areas were reconfigured and recalibrated such as:

• Several interventions were adjusted based on better effectiveness of interventions for certain groups.
• Changes were made in the dosage for each modality of intervention.
• Several policy areas were reviewed and benchmarked for performance measurement.
• Staff planning, workload distribution, staff training and staff performance was linked to outcomes.
• Stakeholders and grant funders were provided evidence of outcome measurements and success.