💰 WiseSpend

Transform Financial Fog Into Actionable Resilience Through AI-Powered Receipt Intelligence

For judges: This README is structured so you can quickly see the problem, solution, architecture, all features (including the Grow page), and a suggested evaluation path. Use Product at a Glance and Demo Scenarios to know what to try in the app; use Evaluation Summary for scoring alignment.

🎯 The Problem

Most people don't realize what they're spending until it's too late.

"I'll just track it manually" → Gets forgotten after 3 days
"I'll use a budgeting app" → Too generic, doesn't understand context
"I already spent the money, why track it?" → Missing the point entirely

The real issue: People don't see the trade-offs of their spending decisions. That $6.50 latte isn't just coffee—it's 15 minutes of your future retirement, or 2 weeks further from your house deposit.

💡 Our Solution

WiseSpend is an agentic, AI-first financial coach that transforms "financial fog" into actionable resilience.

Instead of just tracking expenses, we:

Scan receipts with Gemini 3 → Multimodal AI understands what you bought and why it matters
Calculate real goal impact → Shows you: "That $100 purchase = 2 weeks further from your house deposit"
Socratic coaching → Asks thought-provoking questions, not generic advice
Full observability → Every reasoning step traced in Opik for continuous improvement

The receipt scan isn't about that ONE purchase — it's about building awareness so your NEXT 100 purchases are different.

🏗️ Architecture

System Overview

┌─────────────────────────────────────────────────────────────────────────┐
│                           WISESPEND                                    │
├─────────────────────────────────────────────────────────────────────────┤
│                                                                         │
│  ┌─────────────┐     ┌──────────────┐     ┌─────────────────────────┐  │
│  │   Next.js   │────▶│   Express    │────▶│   Google GenAI Service  │  │
│  │   Frontend  │◀────│   Backend    │◀────│   (Gemini 3)            │  │
│  └─────────────┘     └──────────────┘     └─────────────────────────┘  │
│        │                    │                         │                 │
│        │                    ▼                         ▼                 │
│        │             ┌──────────────┐         ┌─────────────┐          │
│        │             │   MongoDB    │         │   Cloudinary│          │
│        │             │   Database   │         │  (Receipts) │          │
│        │             └──────────────┘         └─────────────┘          │
│        │                                             │                  │
│        │                    └────────────┬───────────┘                  │
│        │                                 ▼                              │
│        │                    ┌─────────────────────────┐                 │
│        └───────────────────▶│         OPIK            │                 │
│           (View Traces)     │  (Full Observability)   │                 │
│                             └─────────────────────────┘                 │
│                                                                         │
└─────────────────────────────────────────────────────────────────────────┘

Receipt Analysis Reasoning Chain (Fully Traced in Opik)

    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │   RECEIPT SCAN   │───▶│  IMAGE ANALYSIS  │───▶│  ITEM EXTRACTION │
    │                  │    │                  │    │                  │
    │ User uploads     │    │ Gemini 3          │    │ Parse line items │
    │ receipt photo    │    │ Vision processes  │    │ Categorize items │
    │                  │    │                  │    │                  │
    └──────────────────┘    └──────────────────┘    └──────────────────┘
                                                              │
                                                              ▼
    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │  GOAL IMPACT     │◀───│  SPENDING PATTERN│◀───│  CONTEXT ANALYSIS│
    │  CALCULATION     │    │  ANALYSIS        │    │                  │
    │                  │    │                  │    │ User history     │
    │ "This $100       │    │ "You usually     │    │ Financial goals  │
    │  purchase = 2    │    │  spend $X on     │    │ Monthly income   │
    │  weeks further   │    │  Fridays"        │    │                  │
    │  from goal"      │    │                  │    │                  │
    └──────────────────┘    └──────────────────┘    └──────────────────┘
                                                              │
                                                              ▼
    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │  SOCRATIC COACH  │───▶│  INSIGHT          │───▶│  USER ACTION     │
    │  QUESTION        │    │  GENERATION       │    │                  │
    │                  │    │                  │    │ User sees        │
    │ "Is this worth   │    │ "15% spike in     │    │ trade-off, makes │
    │  delaying your   │    │  coffee spending  │    │ informed choice  │
    │  house deposit?" │    │  this week"       │    │                  │
    └──────────────────┘    └──────────────────┘    └──────────────────┘
    
    ⬆️ Every step traced in Opik with nested parent-child spans

Grow Page Flow (Knowledge + Investments + Chat)

    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │   USER OPENS      │───▶│  GROW PAGE       │───▶│  TWO TABS        │
    │   GROW            │    │                  │    │                  │
    │   (sidebar)       │    │  Learning &      │    │  Knowledge       │
    │                   │    │  Investing hub   │    │  | Investments   │
    └──────────────────┘    └──────────────────┘    └────────┬─────────┘
                                                              │
           ┌──────────────────────────────────────────────────┼──────────────────────────────────────────────────┐
           │                                                  │                                                  │
           ▼                                                  ▼                                                  ▼
    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │  KNOWLEDGE TAB   │───▶│  LITERACY CARDS  │───▶│  GENERATE QUIZ   │───▶│  GEMINI CURATES  │───▶│  USER TAKES QUIZ │
    │                  │    │  12 topics       │    │  (5 questions)   │    │  5 MC questions  │    │  Score + explain │
    │  50/30/20,       │    │  Click topic     │    │  On demand       │    │  Traced in Opik  │    │  "New quiz"      │
    │  compound, etc.  │    │                  │    │                  │    │                  │    │  anytime         │
    └──────────────────┘    └──────────────────┘    └──────────────────┘    └──────────────────┘    └──────────────────┘
           │
           │
           ▼
    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │  INVESTMENT TAB  │───▶│  ALPHA VANTAGE   │───▶│  CURATED LIST    │
    │                  │    │  Real-time data  │    │  ETFs, stocks    │
    │  Filters: type   │    │  Prices, % change│    │  Risk, min $     │
    │  Risk, sort      │    │                  │    │  Descriptions    │
    └──────────────────┘    └──────────────────┘    └────────┬─────────┘
                                                              │
                                                              ▼
    ┌──────────────────┐    ┌──────────────────┐    ┌──────────────────┐
    │  AI CHAT PANEL   │◀───│  CONTEXTUAL Q&A  │───▶│  USER ACTION     │
    │  (side of Grow)  │    │                  │    │                  │
    │                  │    │  "What is VTI?"  │    │  Informed about  │
    │  Same as AI      │    │  "Explain 50/30" │    │  concepts &      │
    │  Coach; Opik     │    │  Traced in Opik  │    │  opportunities   │
    │  traced          │    │                  │    │                  │
    └──────────────────┘    └──────────────────┘    └──────────────────┘

    ⬆️ Quiz generation and Grow chat traced in Opik (chat-gemini-3, quiz generation spans)

Opik Tracing Architecture

OPIK INTEGRATION
├── Trace Creation ──────────── Every AI call creates a trace
│   ├── chat-gemini-3 ─────────── Chat (AI Coach + Grow panel)
│   ├── analyze-receipt-image ─── Receipt analysis (main feature)
│   └── generate-quizzes ──────── Grow: AI-generated literacy quizzes
│
├── Nested Spans ─────────────── Detailed reasoning chain
│   ├── preprocess-receipt-image ─ Image preprocessing
│   ├── gemini-multimodal-analysis ─ LLM vision analysis
│   ├── calculate-goal-impact ─── Goal impact calculation
│   └── (Grow) quiz curation ───── Quiz generation for Knowledge tab
│
├── Error Handling ───────────── All errors logged to traces
│   └── Error metadata captured ─ Error type, message, stack
│
├── Token Usage Tracking ─────── Full cost observability
│   └── promptTokens, completionTokens, totalTokens
│
└── Graceful Shutdown ─────────── Auto-flush on SIGTERM/SIGINT
    └── Ensures all traces sent before exit

✨ Core Features

📸 Multimodal Receipt Scanning

Snap a photo, understand the impact.

Gemini 3 Vision: Uses Gemini 3's multimodal capabilities to parse unstructured receipt images
Intent Recognition: Distinguishes "necessity" from "luxury" (not just OCR)
Automatic Categorization: Smart categorization of line items
Full Opik Tracing: Every step from image → analysis → insight is traced

🤔 Socratic Financial Coaching

Ask questions, not commands.

Instead of: "You spent too much on coffee"

We ask: "Is that $6.50 latte worth 15 minutes of your future retirement?"

Thought-Provoking Questions: Helps users discover their own financial insights
Goal-Aware: Every question references your actual financial goals
Pattern Recognition: Identifies spending patterns over time
Supportive but Firm: Tone tuned via Opik Agent Optimizer

🎯 Goal Impact Calculation

See the real cost of every purchase.

Real-Time Calculations: Shows how each purchase affects your goals
Visual Trade-offs: "That $100 purchase = 2 weeks further from your house deposit"
Historical Context: Compares current spending to past patterns
Actionable Insights: Not just data, but meaning

🛡️ Safety & Moderation (Opik Evaluation)

Ensures our coach never goes rogue.

ModerationMetric: Prevents suggesting high-risk investments
HallucinationMetric: Ensures accurate balance data
LLM-as-a-Judge: Continuous evaluation of AI responses
Regression Testing: 50+ edge-case financial scenarios in Opik

📊 Financial Dashboard

Wise Score: Your financial resilience score (0-100)
Monthly Spending: Real-time spending tracking
Emergency Fund: Safety net visualization
Transaction History: All scanned receipts with insights
Analytics: Spending patterns, trends, and predictions

🎓 Gamified Financial Learning (AI-Curated)

Learn financial concepts through interactive quizzes.

12 Financial Literacy Topics: Covering budgeting, investing, debt management, and more
- The 50/30/20 Rule
- Compound Growth
- Inflation vs. Cash
- Emergency Fund Basics
- Debt Snowball Method
- Diversification
- Automated Savings
- Tax-Advantaged Accounts
- Credit Score Basics
- ROI Calculation
- Budgeting Apps
- Financial Goals (SMART framework)
Interactive Quiz System:
- Click any literacy card to start a quiz
- Multiple-choice questions with explanations
- Progress tracking and completion rewards
- AI-curated content tailored to your financial profile
Gamification Elements:
- Visual progress indicators
- Completion celebrations
- Topic-based learning paths
- Knowledge reinforcement through repetition

💼 Investment Opportunity Curation

AI-powered investment suggestions based on real market data.

Real-Time Market Data: Powered by Alpha Vantage API
- Live stock prices and ETF quotes
- Daily change percentages
- Market trends and top gainers
Curated Investment Opportunities:
- ETFs: SPY (S&P 500), VTI (Total Stock Market), VEA (International), BND (Bonds), VNQ (Real Estate)
- Major Stocks: Apple, Microsoft, Alphabet, and more
- Risk Assessment: Low, Medium, Medium-High risk classifications
- Minimum Investment: Shows accessible entry points ($1+)
Smart Filtering:
- Filters by investment type (ETF, Stock, Bond, REIT)
- Sorted by performance and risk level
- Educational descriptions for each opportunity
- Sector information and market insights
Safety Features:
- ModerationMetric ensures no high-risk suggestions
- Educational disclaimers
- Encourages consultation with financial advisors
- Past performance disclaimers

🔄 Agent Optimization (Opik Agent Optimizer)

Continuously improving prompts.

Prompt Tuning: System prompts optimized for "supportive but firm" tone
A/B Testing: Different prompt variations tested
Performance Tracking: Completion rates, user satisfaction
Auto-Improvement: Prompts refined based on real user interactions

🔍 Opik Integration (Deep)

WiseSpend showcases production-grade Opik integration with comprehensive tracing, evaluation, and optimization.

Opik integration — files for reviewers (traverse here):

File	Role
`backend/src/infrastructure/services/OpikService.ts`	Core Opik client: init, `createTrace()`, `flush()`, env config (OPIK_API_KEY, OPIK_PROJECT_NAME, OPIK_WORKSPACE)
`backend/src/infrastructure/services/GoogleGenAIService.ts`	Traces for chat and receipt analysis reasoning chain; creates spans (e.g. `analyze-receipt-image`, chat trace), logs token usage and errors
`backend/src/application/use-cases/FinancialAssistantChatUseCase.ts`	Trace `financial-assistant-chat` for AI Coach (and Grow panel) conversations
`backend/src/application/use-cases/GenerateQuizzesUseCase.ts`	Trace `generate-personalized-quizzes` for Grow Knowledge tab quiz generation
`backend/src/infrastructure/services/QuizCuratorAgent.ts`	Nested trace `generate-quiz-for-concept`; flat input/output for Opik LLM-as-a-Judge
`backend/src/infrastructure/services/QuizEvaluator.ts`	Trace `quiz-llm-judge-evaluation`; LLM-as-a-Judge evaluation of quiz content (when EVALUATE_QUIZZES=true)
`backend/src/index.ts`	Graceful shutdown: calls `opikService.flush()` on SIGTERM/SIGINT so traces are sent before exit

📊 Full Tracing Coverage

Every AI operation is traced:

Operation	Trace Name	Spans
Chat Conversation	`chat-gemini-3`	LLM call, response processing
Receipt Analysis	`analyze-receipt-image`	Preprocessing, Vision analysis, Goal impact calculation

Trace Structure:

trace: analyze-receipt-image
├── span: preprocess-receipt-image (type: general)
├── span: gemini-multimodal-analysis (type: llm)
│   ├── input: { model, prompt, image, temperature }
│   ├── output: { analysis, fullAnalysis }
│   └── metadata: { usage: { tokens }, provider: "google-genai" }
└── span: calculate-goal-impact (type: general)
    └── input: { analysis }

🎯 Evaluation Metrics

Built-in Opik Evaluators:

ModerationMetric: Ensures no high-risk investment suggestions
HallucinationMetric: Validates financial data accuracy
Custom Evaluators: Strategy alignment, personalization quality

Evaluation Workflow:

AI Response Generated
        │
        ▼
┌───────────────────────┐
│  Opik Evaluation      │
│  • ModerationMetric    │
│  • HallucinationMetric│
│  • Custom Evaluators  │
└───────────────────────┘
        │
        ▼
┌───────────────────────┐
│  Response Approved    │
│  or Flagged           │
└───────────────────────┘

🔧 Agent Optimization

Opik Agent Optimizer Integration:

System Prompt Tuning: Optimized for "supportive but firm" coaching tone
Performance Tracking: Measures user engagement and goal completion rates
Iterative Improvement: Prompts refined based on real-world performance
Regression Testing: 50+ edge-case scenarios ensure stability

📈 Observability Features

Token Usage Tracking: Full cost observability for every LLM call
Error Tracking: All errors captured with full context
Performance Metrics: Response times, success rates
User Journey Mapping: Complete trace of user interactions

🚀 Quick Start

Prerequisites

Node.js 18+ and npm
MongoDB (local or Atlas)
Google Cloud account (for Gemini API)
Comet account (for Opik)

1. Backend Setup

cd backend

# Install dependencies
npm install

# Copy environment variables
cp .env.example .env

# Edit .env with your keys:
# - GEMINI_API_KEY (from Google Cloud)
# - OPIK_API_KEY (from Comet)
# - OPIK_PROJECT_NAME (e.g., "WiseSpend-Evals")
# - OPIK_WORKSPACE (your Comet workspace)
# - MONGODB_URI (your MongoDB connection string)
# - CLOUDINARY_* (for receipt image storage)

# Start development server
npm run dev

Backend runs on http://localhost:8000

2. Frontend Setup

cd frontend

# Install dependencies
npm install

# Copy environment variables
cp .env.example .env

# Edit .env with your keys:
# - NEXT_PUBLIC_API_URL=http://localhost:8000/api/v1
# - GOOGLE_CLIENT_ID (for OAuth)
# - GOOGLE_CLIENT_SECRET (for OAuth)
# - NEXTAUTH_SECRET (generate with: openssl rand -base64 32)
# - NEXTAUTH_URL=http://localhost:3000

# Start development server
npm run dev

Frontend runs on http://localhost:3000

3. Access the Application

Frontend: http://localhost:3000
Backend API: http://localhost:8000/api/v1
Opik Dashboard: https://www.comet.com/opik

🎬 Demo Scenarios

Scenario 1: First Receipt Scan

User uploads coffee receipt → Gemini analyzes image
System extracts: "Coffee - $6.50"
Goal impact calculated: "That's 15 minutes of your retirement fund"
Socratic question: "Is this worth delaying your house deposit by 2 weeks?"
Full trace in Opik: See the complete reasoning chain

Scenario 2: Pattern Recognition

User scans 5 receipts → System identifies pattern
Insight generated: "You spend 15% more on coffee when stressed"
Coaching question: "Noticed a spike in convenience food spending. What's driving this?"
User reflects → Makes informed choice next time

Scenario 3: Goal Achievement

User sets goal: "Save $10,000 for house deposit"
Scans receipts → Each purchase shows impact
Visual progress: "You're 2 weeks closer to your goal!"
Motivation: Seeing real progress encourages continued saving

Scenario 4: Grow Page — Knowledge + Quiz

User opens Grow (sidebar) → Defaults to Knowledge tab
Sees financial literacy cards → 12 topics (50/30/20, compound growth, emergency fund, etc.)
Clicks a topic → Generate quiz → AI returns 5 multiple-choice questions
Completes quiz → Sees score and explanations; can generate a new quiz anytime

Scenario 5: Grow Page — Investment Discovery

User opens Grow → Switches to Investment Suggestions tab
Sees curated opportunities → ETFs (SPY, VTI, VEA, BND, VNQ), stocks; real-time prices (Alpha Vantage)
Filters by type → ETF, Stock, Bond, REIT; reviews risk levels and descriptions
Uses side AI panel → Asks follow-up questions about a ticker or concept
Makes informed decision → With full context and disclaimers

📋 Evaluation Summary

Criteria	Implementation
✅ True Agent	Autonomous reasoning chain: Receipt → Analysis → Goal Impact → Coaching
⭐ Deep Opik Integration	Every AI call traced with nested spans, full observability
✅ Evaluation-Driven	ModerationMetric, HallucinationMetric, custom evaluators
✅ Agent Optimization	Opik Agent Optimizer tunes prompts for optimal coaching tone
✅ Production Ready	Error handling, graceful shutdown, comprehensive logging
✅ Novel Use Case	Multimodal receipt analysis with goal impact calculation
✅ User Value	Transforms "financial fog" into actionable insights
✅ Gamified Learning	🎓 Interactive financial literacy quizzes (12 topics)
✅ Investment Curation	💼 Real-time investment opportunities with market data
✅ Full Stack	Next.js frontend, Express backend, MongoDB, Cloudinary
✅ Authentication	Google OAuth via NextAuth, protected routes

📁 Key Files

backend/
├── src/
│   ├── infrastructure/
│   │   └── services/
│   │       ├── GoogleGenAIService.ts    # 🤖 Gemini integration with Opik tracing
│   │       └── OpikService.ts           # 🔍 Opik client wrapper
│   ├── application/
│   │   └── use-cases/
│   │       ├── AnalyzeReceiptUseCase.ts  # 📸 Receipt analysis logic
│   │       ├── ChatUseCase.ts            # 💬 Socratic coaching
│   │       └── ComputeWiseScoreUseCase.ts # 📊 Financial resilience score
│   └── presentation/
│       └── controllers/
│           ├── ReceiptController.ts      # Receipt upload & analysis
│           └── AIController.ts           # AI chat endpoints
│
frontend/
├── app/
│   ├── dashboard/
│   │   ├── page.tsx                     # 📊 Main dashboard
│   │   ├── transactions/page.tsx        # 📝 Transaction history
│   │   ├── analytics/page.tsx           # 📈 Spending analytics
│   │   ├── goals/page.tsx               # 🎯 Financial goals
│   │   ├── grow/page.tsx                # 🌱 Grow page — Knowledge tab (literacy + AI quizzes) + Investment Suggestions tab
│   │   └── ai-coach/page.tsx            # 🤔 Socratic coach chat
│   └── scan/page.tsx                    # 📸 Receipt scanning
├── components/
│   ├── dashboard/
│   │   ├── FinancialSummaryCards.tsx   # Summary cards
│   │   ├── SocraticCoach.tsx            # Coach insights
│   │   ├── TransactionList.tsx          # Recent transactions
│   │   ├── FinancialLiteracyCards.tsx   # 🎓 Gamified learning cards
│   │   ├── QuizModal.tsx                # 📝 Interactive quiz component
│   │   ├── InvestmentSuggestions.tsx    # 💼 Investment opportunities
│   │   └── FinancialInsights.tsx        # 💡 AI-generated insights
│   └── receipt/
│       └── AnalysisResultModal.tsx      # Receipt analysis results
└── services/
    ├── receiptService.ts                 # Receipt API calls
    ├── aiService.ts                      # AI chat API calls
    └── investmentService.ts              # 💼 Investment data (Alpha Vantage)

🔧 Technology Stack

Frontend

Next.js 14 - React framework with App Router
TypeScript - Type safety
Tailwind CSS - Styling
NextAuth - Authentication
Framer Motion - Animations

Backend

Express.js - REST API server
TypeScript - Type safety
MongoDB - Database (via Mongoose)
Cloudinary - Receipt image storage

AI & Observability

Google Gemini 3 - Multimodal AI (vision + text)
Opik - Full observability and evaluation
Opik Agent Optimizer - Prompt optimization

📊 Opik Dashboard

View your traces:

Go to https://www.comet.com/opik
Navigate to your workspace
Open project: WiseSpend-Evals (or your configured project)
See traces:
- chat-gemini-3 - Chat conversations
- analyze-receipt-image - Receipt analysis (with nested spans)

Trace Details:

Click any trace to see:
- Full input/output
- Nested spans (preprocessing → analysis → goal impact)
- Token usage
- Timing information
- Error logs (if any)

🧪 Testing Opik Integration

Test Receipt Analysis

# Start backend
cd backend && npm run dev

# In another terminal, test receipt analysis
curl -X POST http://localhost:8000/api/v1/ai/receipt/analyze \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -F "image=@/path/to/receipt.jpg"

Test Chat Endpoint

curl -X POST http://localhost:8000/api/v1/ai/chat \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "What is the capital of France?"}
    ],
    "temperature": 0.7,
    "maxTokens": 100
  }'

View Traces in Opik

After making API calls:

Wait 10-30 seconds (traces are batched)
Go to Opik dashboard
Refresh and see your traces!

💡 Philosophy

"The receipt scan isn't about that ONE purchase — it's about building awareness so your NEXT 100 purchases are different."

WiseSpend transforms hindsight into foresight. By showing you the real cost of every purchase, you start making informed choices before you spend, not after.

📝 License

Built for the Comet "Commit to Change" AI Agents Hackathon

🙏 Acknowledgments

Google Gemini - Multimodal AI capabilities
Comet Opik - Observability and evaluation platform
Comet Team - Hackathon organizers and support

Built with 💰 for the Comet AI Agents Hackathon

Name		Name	Last commit message	Last commit date
Latest commit History 53 Commits
backend		backend
frontend		frontend
.DS_Store		.DS_Store
OPIK_DROPDOWN_FIX.md		OPIK_DROPDOWN_FIX.md
OPIK_LLM_JUDGE_SETUP.md		OPIK_LLM_JUDGE_SETUP.md
PITCH_DECK.md		PITCH_DECK.md
README.md		README.md

Folders and files

Latest commit

History

Repository files navigation

💰 WiseSpend

🎯 The Problem

💡 Our Solution

🏗️ Architecture

System Overview

Receipt Analysis Reasoning Chain (Fully Traced in Opik)

Grow Page Flow (Knowledge + Investments + Chat)

Opik Tracing Architecture

✨ Core Features

📸 Multimodal Receipt Scanning

🤔 Socratic Financial Coaching

🎯 Goal Impact Calculation

🛡️ Safety & Moderation (Opik Evaluation)

📊 Financial Dashboard

🎓 Gamified Financial Learning (AI-Curated)

💼 Investment Opportunity Curation

🔄 Agent Optimization (Opik Agent Optimizer)

🔍 Opik Integration (Deep)

📊 Full Tracing Coverage

🎯 Evaluation Metrics

🔧 Agent Optimization

📈 Observability Features

🚀 Quick Start

Prerequisites

1. Backend Setup

2. Frontend Setup

3. Access the Application

🎬 Demo Scenarios

Scenario 1: First Receipt Scan

Scenario 2: Pattern Recognition

Scenario 3: Goal Achievement

Scenario 4: Grow Page — Knowledge + Quiz

Scenario 5: Grow Page — Investment Discovery

📋 Evaluation Summary

📁 Key Files

🔧 Technology Stack

Frontend

Backend

AI & Observability

📊 Opik Dashboard

🧪 Testing Opik Integration

Test Receipt Analysis

Test Chat Endpoint

View Traces in Opik

💡 Philosophy

📝 License

🙏 Acknowledgments

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages