Realistic, insurance-relevant challenges for high-potential technology talent.

Catastrophe response requires rapid integration of weather, hazard, geographic, infrastructure and community data into an explainable operational picture. Public data is abundant, but formats, spatial resolution, timing and quality vary significantly. This track incorporates the Geospatial Catastrophe Impact Intelligence concept while also accommodating AI-driven damage assessment capabilities.

• Ingest public weather, disaster, hazard and catastrophe-event data.
• Normalize coordinates, geographic identifiers, spatial resolution and event timestamps.
• Generate and visualize hazard footprints and potentially affected regions.
• Integrate organizer-approved satellite or aerial imagery and associated metadata.
• Apply computer vision or multimodal methods to identify visible impact indicators and estimate relative damage severity.
• Correlate event, location, imagery, infrastructure and community-level signals.
• Track changes in event footprint and severity over time.
• Explain the data and visual signals behind each prioritization or severity assessment.
• Provide performance maps, dashboards, alerts and versioned APIs.
• Communicate imagery limitations, uncertainty, resolution, coverage and human-review requirements.

Public cyber information is fragmented across vulnerability databases, known-exploitation catalogues, vendor advisories and news. Severity alone does not indicate systemic relevance, and product names and versions are frequently inconsistent across sources. This track combines the Cyber Risk Signal & Systemic Technology Intelligence concept with the GenAI Business Intelligence Agent idea into a single, broader intelligence and decision-support theme.

• Ingest and normalize structured and unstructured data.
• Resolve vendor, product, technology and other relevant entity aliases across approved sources.
• Build and query relationships using knowledge graphs or equivalent explainable relationship models.
• Identify emerging risk signals and material changes over time.
• Analyze multi-source evidence, including conflicting or incomplete information.
• Implement retrieval-augmented generation and source-grounded natural-language generation.
• Generate actionable insights, briefs, watchlists, timelines and recommended Next Best Actions.
• Provide a conversational layer that supports business questions, follow-up questions and evidence retrieval.
• Support agentic workflows with explicit boundaries, human approval and auditable steps.
• Provide full-stack dashboards, APIs and traceable user experiences.