Team Name:
TrustAid
Team Members:
TrustAid
TrustAid helps people find and act on the right government support in minutes. It’s built for real life events—bereavement, becoming a carer, job loss—providing personalised guidance, local service connections, and plain-English checklists. Every answer is evidence-first: TrustAid cites official pages and open datasets, shows dates, and for data questions runs read-only queries with the exact SQL and an audit_id for full traceability. If strong evidence isn’t found, it won’t guess—it flags the gap and suggests the next check. Privacy is in the user’s control: keep chat history or use a temporary session; data stays private and read-only. The result: reliable aid, with clear, actionable next steps.
Description of Use Powers TrustAid’s explainability layer: we reference these statements in the UI (citations, confidence, audit_id, model mode) and use them to justify refusals/guardrails in sensitive scenarios.
Description of Use Provides program/spend context behind guidance. TrustAid links to entity reports as evidence, and can show high-level spend/performance facts alongside service recommendations to improve trust.
Description of Use Acts as TrustAid’s ingestion index—we discover and refresh source tables/documents from here, surface dataset provenance in answers, and let users follow through to the original dataset page.
Description of Use TrustAid prioritises and badges compliant services in recommendations, and cites compliance status to explain why a particular channel or service is preferred (transparency + quality).
Description of Use Helps TrustAid set expectations and route users—e.g., flag degraded services, suggest alternative channels/times, and display performance badges next to links users are about to visit.
Go to Challenge | 18 teams have entered this challenge.
Go to Challenge | 34 teams have entered this challenge.
Go to Challenge | 27 teams have entered this challenge.
Go to Challenge | 13 teams have entered this challenge.
