Project Document — Full Detail·Engagement
Prepared for a cultural & academic district partnership · Version 1.0 · Proof of Concept · June 2026
A privacy-first, institution-autonomous intelligence layer that enables personalized, coordinated engagement across the district — without replacing existing institutional systems.
District Connect builds a secure, shared data spine that unifies engagement history, event calendars, and donor insights across anchor institutions — a leading art museum, a research university, and a premier music center — without replacing existing institutional CRMs. Engagement across the district is fragmented today: each institution operates its own systems, calendars, and outreach in isolation. District Connect moves partners from isolated transaction models to a shared relationship-intelligence network, where a single supporter's relationship with the district is visible, coordinated, and stewarded as a whole.
In Scope
Cross-institution event discovery, donor segmentation, coordinated campaigns, consent management, and the public Community Concierge experience.
Out of Scope
AI/ML predictive models, real-time data sync, native mobile applications, and enterprise-scale infrastructure — all candidates for post-POC phases.
Demonstrate the value of cross-institution data sharing — with privacy preserved throughout.
Validate the technical feasibility of a multi-tenant, institution-autonomous architecture.
Establish a durable governance framework for data sharing across independent institutions.
Measure the engagement lift produced by coordinated, consent-based outreach.
Activation — Front-facing experience
Personalized, multi-stop cross-district itineraries. Discover the Circle Portal — one interface to search programs by interest or accessibility. Campaign Studio — align event timing, resolve overlaps, run cooperative outreach.
Intelligence — Cross-institution insights
Securely unifies membership, volunteer, and ticketing behavior. Next-Best-Action Engine — pre-validated, automated next actions. Shared Affinity & Overlap — affinity scores and segment-overlap flags across CRMs.
Foundation — Trust & governance
Permanent, auditable opt-in records. Staff-Review Mandate — a draft-generation engine only, never autonomous outreach. Governed Spine — privacy-by-design architecture with immutable, reviewable logs.
A non-intrusive intelligence layer that sits above each institution's existing operational infrastructure — without replacing it.
Structured batch processing (CSV/JSON via secure SFTP) for a low-friction POC; scales to live REST APIs post-pilot.
Salted SHA-256 email hashes with name/zip confidence scoring — resolves duplicates without reading raw identities.
Managed PostgreSQL on Amazon RDS / Aurora Serverless v2 with row-level security, KMS encryption, automated backups.
Next.js frontend and Python (FastAPI) services on AWS Fargate, behind Amazon CloudFront with AWS WAF.
Critical Infrastructure Rule
Raw personally identifiable information is isolated and tokenized at the institutional perimeter. The central spine retains only anonymized IDs, salted email hashes, interest tags, and explicit consent vectors — never raw identity.
TLS 1.3 for all data in transit. AES-256 for data at rest.
Institution admins own source records only. Coordinators get read-only aggregated metrics. Auditors get read-only immutable logs.
90-day deletion after project close, with certificates of destruction. Consent logs retained 7 years for compliance.
Runs on a cloud-native AWS stack — est. ~$300 infrastructure cost, largely offset by AWS credits.