Evidentia

• 

  Track 3 : Copilot

• 

  Explainability

• 

  Home Page - Command Center

• 

  Track 2 : Desert Planner

• 

  Trust Fabric

• 

  Signal Intergrity

• 

  Planning Studio

• 

  Expert Validation

• 

Inspiration

Healthcare systems don’t suffer from a lack of data—they suffer from a lack of trusted decisions.Infrastructure is often built where data exists, not where real need exists.

We were inspired to solve a deeper problem:How do you help decision makers trust messy, incomplete, and conflicting healthcare data?

What it does

Evidentia is an evidence-aware healthcare planning intelligence platform that helps:

• Identify true gaps in healthcare access

• Distinguish real shortages vs. missing data

• Recommend trusted facilities for patient care

• Provide explainable, evidence-backed decisions

• Enable planners to save, review, and refine decisions

Supports:

• Medical Desert Planner (Track 2) — Find gaps with confidence

• Referral Copilot (Track 3) — Recommend where patients should go

How we built it

Built on Databricks Lakehouse, Streamlit, FIASS,

• Data Pipeline - Cle

• Data Processing

  • Unified 10,000 messy facility records
  • Combined structured and unstructured data

• Evidence Extraction

  • Converted free text into verifiable capability signals
  • Generated confidence scores

• Signal Integrity Engine

  • Differentiates:

    • Real infrastructure gaps
    • Data quality issues

• Models

  • Geospatial coverage analysis (Track 2)
  • Distance + trust-based ranking (Track 3)

• App Layer

  • Simple non-technical workflows
  • Evidence citations for every output
  • Scenario saving and overrides

Challenges we ran into

• Messy data

  • Incomplete and inconsistent facility claims

• Signal vs. noise

  • Hard to distinguish real gaps from missing data

• Explainability

  • Every output needed clear supporting evidence

• Human trust

  • Avoiding overconfidence required confidence scoring

Accomplishments that we're proud of

• Treated data as claims, not ground truth

• Built a Signal Integrity Engine for trust-aware decisions

• Delivered explainable recommendations with citations

• Enabled confidence-aware planning, not just analytics

• Designed for non-technical users

What we learned

• Trust matters more than data

• Uncertainty must be visible

• Explainability is critical in healthcare

• Human-in-the-loop is essential

What's next for Evidentia

• Integrate real-time healthcare data

• Add predictive demand modeling

• Improve geospatial simulations

• Expand to multi-country deployments

• Enable collaborative planning workflow

Built With

• databricks
• fiass
• streamlit