DonorConnect

Making a difference together

Evidence & Rubric Alignment

Clear mapping of competencies to implementation evidence with direct links.

CCC.1.3CCC.1.4CCC.1.5TS.6.2TS.6.3
CCC.1.3

Implement a Solution

"I can implement a solution using the most appropriate industry-accepted method that leads to effective implementation of the solution."

📋 Rubric Requirements

  • ✓ Build a working MVP with multiple pages
  • ✓ Use industry-accepted methods (Agile, DevOps, Interface design)
  • ✓ Apply appropriate tools and best practices

🎯 How I Demonstrated This

Working MVP: 12+ Functional Pages

Dashboard AnalyticsReal-time metrics & KPIs/dashboard →Donor ManagementFull CRUD operations/donors →Add/Edit DonorForm validation/donors/new →Donation TrackingRecord & list gifts/donations →Record DonationCampaign linking/donations/new →Campaign ManagementGoal tracking/campaigns →Segment BuilderDynamic grouping/segments →Workflow AutomationMulti-step sequences/workflows →Task ManagementFollow-up tracking/tasks →Login PageSecure authentication/login →RegistrationOrg creation/register →AI PolicyDocumentation/ai-policy →

Industry-Accepted Methods & Tools

Development Methodology

  • • Agile/Iterative – Built features incrementally
  • • Component-based – Reusable React components
  • • RESTful API – Standard HTTP endpoints
  • • Git version control – GitHub repository

Technology Stack

  • • Next.js 16 – Production React framework
  • • PostgreSQL + Prisma – Industry-standard database
  • • Tailwind CSS – Modern styling approach
  • • Vercel – CI/CD deployment

📎 Direct Evidence Links

📂 GitHub Source Code🌐 Live Deployed App📝 Notion Docs (Labs, Wireframes, Trello)📊 View Dashboard
CCC.1.4

Test and Improve a Solution

"I can conduct extensive testing of my solution with users, evaluate my design, and identify ways to improve its effectiveness in solving the problem."

📋 Rubric Requirements

  • ✓ Test my solution with users and identify areas for improvement
  • ✓ Gather targeted feedback about my solution
  • ✓ Use feedback to revise and refine my solution

🎯 How I Demonstrated This

🔍 User Testing & Feedback

  • Manual testing of all CRUD flows✓ Complete
  • Cross-browser compatibility check✓ Chrome, Firefox
  • UI responsiveness testing✓ Desktop, Mobile
  • Error handling verification✓ API responses

🔧 Issues Fixed Based on Testing

  1. 1Fixed dark text on dark background (dropdowns)
  2. 2Replaced text inputs with proper select menus
  3. 3Added duplicate email prevention for donors
  4. 4Fixed campaign deletion (unlink donations first)
  5. 5Added working filters to donations page

📝 Reflection on Improvements

Documented challenges, what I learned, and what I would do differently:

View Full Reflection → /reflection
CCC.1.5

Document and Present

"I can document my solution clearly and present it effectively to stakeholders."

📋 Rubric Requirements

  • ✓ Clear documentation of the solution
  • ✓ Explain technical decisions and architecture
  • ✓ Present the solution effectively

🎯 How I Demonstrated This

📖 README.md

Comprehensive project overview, setup instructions, architecture explanation

View on GitHub →

🤖 AI Policy Page

Detailed AI integration documentation, ethics, safeguards

/ai-policy →

💭 Reflection Page

Challenges, learnings, growth throughout the project

/reflection →

📄 In-App Documentation

  • ✓ Code comments explaining complex logic
  • ✓ JSDoc annotations on key functions
  • ✓ Zod schemas documenting data structures
  • ✓ Clear component/file naming conventions

🗄️ Entity Relationship Diagram (ERD)

Complete database schema showing all entities, attributes, and relationships:

This Mermaid ERD diagram documents all 12 entities, their attributes with types, and relationships with cardinality notation. The schema is implemented in prisma/schema.prisma.

TS.6.2

Use AI Responsibly

"I can consider bias, ethics, security, and data privacy when using and building AI systems."

📋 Rubric Requirements

  • ✓ Consider bias in AI outputs
  • ✓ Address ethics of AI usage
  • ✓ Implement security measures
  • ✓ Protect data privacy

🎯 How I Demonstrated This

🔒 Data Privacy

  • ✓Minimal data sent to AI – Only aggregated metrics (gift counts, totals), never raw PII like addresses or phone numbers
  • ✓Organization scoping – Users can only access their own org's donor data
  • ✓No data retention – AI requests are transient, not logged

🛡️ Security

  • ✓Server-side only – API keys stored in env vars, never exposed to client
  • ✓Authentication required – All AI endpoints require valid session
  • ✓Error handling – Failures return safe messages, no stack traces

⚖️ Ethics

  • ✓AI assists, doesn't decide – Provides suggestions for staff to act on, not automated decisions
  • ✓User opt-in – AI feature activated by explicit button click
  • ✓Transparent documentation – AI Policy page explains how AI is used

⚠️ Bias Mitigation

  • ✓Factual prompts – "Do not invent data" instruction reduces hallucinations
  • ✓Low temperature – 0.4 setting prioritizes consistency over creativity
  • ✓Data-driven only – AI summarizes existing data, doesn't make predictions

📍 Code Evidence

AI endpoint with all safeguards implemented:

src/app/api/ai/summarize-donor/route.js
📖 View Full AI Policy Page → /ai-policy
TS.6.3

Prompt Effectively / Integrate AI Tools

"I can craft specific and targeted prompts for machines to follow in order to generate my desired response."

📋 Rubric Requirements

  • ✓ Craft specific, targeted prompts
  • ✓ Generate desired AI responses
  • ✓ Integrate AI meaningfully into the product

🎯 How I Demonstrated This

🤖 AI Feature: Donor Summary Generator

GPT-4o-mini

AI Model Used

80-100 words

Targeted Output

1-click

User Activation

📝 My Prompt Engineering Strategy

SYSTEM PROMPT

"Be concise, factual, and actionable. Do not invent data."

→ Prevents hallucinations and keeps responses focused

USER PROMPT

"You are a donor success assistant. Summarize this donor in 80-100 words max with risk assessment and suggested next action. Use any notes about the donor to personalize the recommendation."

→ Specific word limit, clear output structure, personalization instruction

Why These Prompts Work:
  • 1.Word limit – Prevents verbose, unfocused responses
  • 2."Do not invent" – Reduces hallucination risk
  • 3.Low temp (0.4) – Consistent, reliable outputs
  • 4.Role assignment – "donor success assistant" provides context
  • 5.Structured output – Risk + action = actionable result
  • 6.Personalization – Uses staff notes for better recommendations

💡 How AI Improves the Solution

❌ Without AI

  • • Staff manually reviews 10+ donation records
  • • Risk assessment is subjective, inconsistent
  • • Personalized outreach takes 5-10 min/donor
  • • New staff lack context on donor history

✅ With AI

  • • Instant summary in one click (<2 seconds)
  • • Consistent, data-driven risk assessment
  • • AI suggests specific next action
  • • Anyone can quickly understand donor context

📍 Implementation Evidence

src/app/api/ai/summarize-donor/route.jsSee detailed prompt documentation on AI Policy page →

🔗 Quick Access - All Evidence Links

📂 GitHub🌐 Vercel📝 Notion Docs📊 Dashboard🤖 AI Policy💭 Reflection