AI Operating System for Executive Decision Making
๐ Built for Google GenAI Academy Hackathon 2026
๐ฅ Watch Demo ย ยทย ๐ Live App ย ยทย ๐ Presentation ย ยทย ๐ Documentation
| โก Intelligent Dataset Analysis | ๐ง Google Gemini Powered | ๐ 8 AI-Driven Modules | ๐ Automated Executive Reports | โ๏ธ Cloud Run Ready |
|---|---|---|---|---|
| Profiles industry, KPIs, outliers | Live AI reasoning + offline fallback | End-to-end decision intelligence | Boardroom-ready in one click | Production Docker + Nginx |
- Overview
- Why SynapseIQ?
- Key Features
- Architecture
- Technology Stack
- Project Structure
- Screenshots
- Demo
- Getting Started
- Environment Variables
- Deployment
- Roadmap
- Team
- License
SynapseIQ is an AI-powered Decision Intelligence Platform that transforms structured business datasets into executive-ready insights, strategic recommendations, interactive business relationship maps, forecasting simulations, and boardroom-ready reports โ all powered by Google Gemini.
Unlike traditional BI dashboards that passively display charts, SynapseIQ acts as a living AI operating system that proactively surfaces strategic narratives, detects anomalies, simulates business scenarios, and guides executives toward high-confidence decisions.
"This isn't another analytics tool. This is an AI strategist embedded inside your business data."
| Capability | Traditional Business Intelligence | โก SynapseIQ |
|---|---|---|
| Output | Static charts and dashboards | Executive decisions and strategic recommendations |
| Analysis | Shows what happened | Explains why it happened + what to do next |
| Reporting | Manual, hours-long process | Auto-generated boardroom reports in seconds |
| AI Integration | Generic chatbot widget | Context-aware strategist with full dataset awareness |
| Insights | Reactive โ discovered after damage | Proactive โ AI Initiative Cards surface issues first |
| Scenarios | Not possible | Live scenario simulation with real-time trade-off analysis |
| Relationships | Siloed metric views | Interactive correlation graph (Strategy Canvas) |
| Confidence | No quality signals | Dynamic AI confidence scoring per dataset |
| Data Limits | Silent failures | Honest AI declarations when data is insufficient |
| Deployment | Enterprise software license | Open-source, Docker-ready, Cloud Run in minutes |
|
AI-narrated business health summary with today's strategic priorities, top opportunities, critical risks, and dynamic confidence scoring โ built for immediate boardroom consumption. |
Interactive React Flow relationship graph that maps mathematical correlations between Revenue, Profit, Marketing ROI, Inventory, and Customer Satisfaction. Hover any node for correlation strength and strategic insight. |
|
Real-time anomaly telemetry matrix detecting meaningful business events โ revenue-profit divergence, satisfaction drops, inventory pressure, and marketing ROI outliers โ with McKinsey-style explanations. |
Gemini-powered split-pane consulting interface delivering structured executive responses: Summary โ Evidence โ Interpretation โ Recommendation โ Confidence โ Follow-up. |
|
Six-lever scenario simulator โ Marketing, Pricing, Inventory, Hiring, Retention, Costs โ with real-time AI trade-off analysis and projected revenue, margin, and risk scoring. |
Chronological narrative ledger translating raw dataset records into a human-readable business story with category filtering and trend visualization. |
|
Seven-section auto-generated executive report: Executive Summary ยท Business Health ยท Key Opportunities ยท Critical Risks ยท Forecast ยท Strategic Recommendations ยท 90-Day Action Plan. |
Upload any CSV business dataset. SynapseIQ auto-profiles the data: detects industry type, infers key business metrics, identifies missing values and Z-score outliers, and adapts all modules to your context. |
Additional Capabilities:
- ๐ฌ AI Narrative Engine โ Full-screen animated narrative post-analysis guiding executives through key findings
- โจ Presentation Mode โ One-click projector-optimized interface with expanded spacing and focused layouts
- ๐งญ Guided Demo Tour โ Automated 3-minute walkthrough built for hackathon judges โ no setup required
- ๐ Honest AI Declarations โ Clear diagnostic banners when data is insufficient; no hallucinated answers
- ๐ Intelligence Mesh โ Ambient animated particle background responding to data state changes
graph TD
A["๐ Dataset Upload\nCSV Drag-and-Drop"] --> B["๐ฌ csvParser.ts\nData Processing Engine"]
B --> C["๐ Statistical Analysis\nKPIs ยท Pearson Correlations\nZ-Score Outliers ยท Missing Values"]
B --> D["๐ญ Auto-Profiling\nIndustry ยท Metrics ยท Timeframe"]
C --> E["๐ค geminiService.ts\nGoogle Gemini API"]
D --> E
E --> F["๐ง Decision Intelligence Engine\nlocalAnalysis.ts"]
F --> G["๐พ store.ts\nZustand Global State"]
G --> H["๐ Executive Brief\nHealth ยท Priorities ยท Confidence"]
G --> I["๐ธ Strategy Canvas\nCorrelation Relationship Graph"]
G --> J["๐ค AI Strategist\nContext Chat ยท Evidence ยท Actions"]
G --> K["๐ Forecast Studio\nScenario Sliders ยท Trade-offs"]
G --> L["๐ Boardroom Report\n7-Section Auto-Report"]
G --> M["๐ก Business Signals\nAnomaly Detection Cards"]
G --> N["๐
Business Timeline\nChronological Narrative"]
style A fill:#0D1117,stroke:#83D18B,stroke-width:2px,color:#fff
style E fill:#0D1117,stroke:#4285F4,stroke-width:2px,color:#fff
style G fill:#0D1117,stroke:#A855F7,stroke-width:2px,color:#fff
The CSV parser computes Pearson correlations, Z-score outliers (ยฑ3ฯ), and missing value ratios. This statistical profile feeds Google Gemini to generate a context-rich analysis JSON, which hydrates the Zustand global store powering every module simultaneously.
sequenceDiagram
participant E as Executive
participant SC as Strategy Canvas
participant S as Zustand Store
participant G as Google Gemini
participant L as Local Reasoning Engine
E->>SC: Hover node (Revenue)
SC->>E: Show correlated nodes + correlation strength
E->>SC: Click node (Revenue)
SC->>S: setActiveNodeId("revenue")
S->>E: Refresh context panel with node data
E->>G: "Why did revenue increase?"
alt Gemini API configured
G->>E: Summary ยท Evidence ยท Recommendation ยท Confidence
else No API key / Offline
L->>E: Dataset-grounded heuristic response
end
Note over E,L: Format: Summary โ Evidence โ Interpretation โ Recommendation โ Confidence โ Follow-up
Every AI response โ whether from Gemini or the local fallback โ follows the same structured executive format. The local reasoning engine uses computed KPI stats (mean, max, min, total) from the parsed dataset to provide data-grounded answers without an API key.
graph LR
Dev["๐ป Local Development<br/>React + TypeScript + Vite"] -->|git push| GH["๐ GitHub Repository<br/>pravalika2307/SynapseIQ"]
GH -->|Automatic CI/CD| Vercel["โฒ Vercel Platform<br/>Build & Deploy"]
Vercel -->|Production Deployment| Web["๐ SynapseIQ Web Application<br/>HTTPS ยท Global CDN"]
Web -->|API Request| Gemini["๐ค Google Gemini API<br/>AI Decision Intelligence"]
Gemini -->|AI Responses| Web
Users["๐จโ๐ผ Executives<br/>๐ Analysts<br/>๐ Managers"] -->|Access via Browser| Web
style GH fill:#0D1117,stroke:#fff,stroke-width:1px,color:#fff
style CR fill:#0D1117,stroke:#4285F4,stroke-width:2px,color:#fff
The multi-stage Dockerfile compiles the React SPA in a Node.js builder container, then serves static assets via a minimal Nginx Alpine image on port 8080 โ ideal for Google Cloud Run's serverless container model.
| Technology | Version | Purpose |
|---|---|---|
| React | 19.x | UI Component Framework |
| TypeScript | 6.x Strict | Type Safety & Developer Experience |
| Vite | 8.x | Build Tool & Dev Server |
| Tailwind CSS | v4 | Utility-first Dark Theme Styling |
| Framer Motion | 12.x | Animations & Page Transitions |
| React Router | v7 | Client-side Hash Routing |
| Zustand | 5.x | Global Application State |
| Technology | Purpose |
|---|---|
| Google Gemini API | Executive briefs ยท Copilot chat ยท Scenario reasoning |
| Local Reasoning Engine | Offline heuristic analysis grounded in computed dataset statistics |
| Pearson Correlation Engine | Business metric relationship mapping for Strategy Canvas |
| Z-Score Outlier Detection | Statistical anomaly identification (ยฑ3ฯ threshold) |
| Dynamic Confidence Scoring | AI quality index based on dataset completeness and outlier density |
| Technology | Version | Purpose |
|---|---|---|
| React Flow (xyflow) | 12.x | Interactive Strategy Canvas node graph |
| Recharts | 3.x | Business charts, scatter plots, line charts |
| Technology | Purpose |
|---|---|
| Docker | Multi-stage production container (Node builder โ Nginx Alpine) |
| Nginx | Static SPA serving with hash-router fallback rules |
| Google Cloud Run | Serverless HTTPS container deployment |
| Google Cloud Build | CI/CD from GitHub to Artifact Registry |
SynapseIQ/
โ
โโโ Dockerfile # Multi-stage production container
โโโ nginx.conf # SPA routing + static serving config
โโโ package.json # NPM manifest (v1.0.0)
โโโ tsconfig.json # TypeScript strict configuration
โโโ vite.config.ts # Vite build configuration
โโโ CHANGELOG.md # Version history
โโโ CONTRIBUTING.md # Contribution guidelines
โโโ SECURITY.md # Security policy
โ
โโโ docs/
โ โโโ screenshots/ # README screenshots (7 pages)
โ
โโโ public/
โ โโโ favicon.svg # Application icon
โ
โโโ src/
โโโ App.tsx # Root HashRouter + route definitions
โโโ main.tsx # React DOM entry point
โโโ index.css # Global design tokens + Tailwind
โ
โโโ components/ # Shared UI components
โ โโโ ui/ # Atomic design system (Button, Card, Badgeโฆ)
โ โโโ DecisionGraph.tsx # React Flow interactive relationship graph
โ โโโ AIMissionControl.tsx # Animated AI narrative overlay
โ โโโ DemoController.tsx # Guided demo tour state machine
โ โโโ IntelligenceMesh.tsx # Ambient animated particle background
โ โโโ LiveInsightStream.tsx # Live activity ticker sidebar
โ โโโ PresentationToolbar.tsx # Presentation mode toggle
โ โโโ Sidebar.tsx # Collapsible navigation panel
โ โโโ Topbar.tsx # Top bar + AI status indicators
โ โโโ UploadZone.tsx # CSV drag-and-drop upload
โ
โโโ features/ # Business logic & AI services
โ โโโ csvParser.ts # CSV ingestion ยท profiling ยท correlation engine
โ โโโ data.ts # Node context and telemetry data maps
โ โโโ defaultDataset.ts # Built-in demo dataset (NovaRetail Q2)
โ โโโ demoStore.ts # Guided tour Zustand state machine
โ โโโ geminiService.ts # Google Gemini API integration
โ โโโ localAnalysis.ts # Offline heuristic analysis engine
โ โโโ store.ts # Global application state (Zustand)
โ
โโโ layouts/
โ โโโ DashboardLayout.tsx # Sidebar + Topbar shell wrapper
โ
โโโ pages/ # Full-page view components
โโโ Landing.tsx # Upload portal + AI intake sequence
โโโ ExecutiveBrief.tsx # Executive summary + health + priorities
โโโ BusinessTimeline.tsx # Chronological event narrative
โโโ BusinessSignals.tsx # Anomaly telemetry matrix
โโโ StrategyCanvas.tsx # Strategy graph + scatter analysis
โโโ DecisionCopilot.tsx # AI strategist chat interface
โโโ Forecast.tsx # Scenario simulator + projections
โโโ Reports.tsx # Boardroom report generator
โโโ DataExplorer.tsx # Raw dataset exploration viewer
All screenshots below show SynapseIQ running live with the built-in NovaRetail Q2 demo dataset. No API key is required for the demo experience.
Landing Portal โ Intelligent Dataset Intake
Drag-and-drop CSV upload ยท Auto-profiling ยท Industry detection ยท Validation feedback
Executive Brief โ AI-Narrated Business Summary
Business Health Score 90/100 ยท Decision Readiness 87/100 ยท AI Confidence 95% ยท Strategic Priorities ยท Live Telemetry
Strategy Canvas โ Interactive Business Relationship Graph
Pearson correlation mapping ยท Revenue $246.78M ยท Node relationship tooltips ยท Animated edges
AI Strategist โ Context-Aware Decision Copilot
Gemini-powered executive chat ยท Dataset Profile panel ยท McKinsey response format ยท Confidence scoring
Forecast Studio โ Live Scenario Simulator
Six business levers ยท Projected Revenue $49.6M ยท Gross Margin 45.7% ยท Real-time AI trade-off analysis
Business Timeline โ Chronological Narrative Ledger
Event-driven business story ยท Category filter chips ยท Trend visualization ยท Seasonal dynamics
Boardroom Report โ Auto-Generated Executive Report
7-section executive brief ยท Strategic Planning dossier ยท Investor-ready format
| Resource | Link |
|---|---|
| ๐ฅ Demo Video | Coming Soon โ Upload to YouTube |
| ๐ Live Deployment | Deploy to Vercel โ see instructions below |
| ๐ Presentation Deck | Google Slides presentation link |
SynapseIQ includes a built-in Guided Demo Tour that requires zero configuration.
- Open the application โ the built-in NovaRetail Q2 dataset loads automatically
- Click Start Guided Demo on the landing screen
- The system auto-navigates through all 8 key features with spotlights and narration
- Completes in under 3 minutes
No Gemini API key required. The guided demo uses an intelligent offline reasoning engine backed by pre-computed statistical analysis of the bundled dataset.
| Requirement | Version |
|---|---|
| Node.js | โฅ 20.x |
| npm | โฅ 10.x |
| Gemini API Key | Optional |
# 1. Clone the repository
git clone https://github.com/pravalika2307/SynapseIQ.git
cd SynapseIQ
# 2. Install dependencies
npm install
# 3. Start the development server
npm run devOpen https://synapseiq-ai.vercel.app/ in your browser.
# Type-check + compile production bundle
npm run build
# Preview the production build locally
npm run preview
# Run ESLint code quality checks
npm run lintSynapseIQ works fully offline with its built-in reasoning engine. A Gemini API key unlocks live AI capabilities.
| Variable | Required | Description |
|---|---|---|
GEMINI_API_KEY |
Optional | Google Gemini API key for live executive briefs and copilot chat |
How to configure: No .env file is required. Enter your Gemini API key directly in the app UI at the landing screen. The key is stored in browser session memory only and is never persisted or transmitted to any external server other than the Google Gemini API.
Privacy: All data processing happens client-side in your browser. No dataset or business data leaves your device unless you explicitly enable Gemini API integration.
SynapseIQ is deployed on Vercel with automatic CI/CD integration.
# Install Vercel CLI
npm install -g vercel
# Login
vercel login
# Deploy
vercel --prodEvery push to the main branch automatically:
- Builds the React + Vite application
- Optimizes production assets
- Deploys to Vercel's global CDN
- Generates a secure HTTPS production URL
docker build -t synapseiq:latest .
docker run -p 8080:8080 synapseiq:latestnpm run build
# Deploy the dist/ directory โ no server-side requirements| Status | Version | Milestone |
|---|---|---|
| โ Complete | v1.0.0 | Hackathon Prototype โ Full AI decision platform with Gemini |
| ๐ In Progress | v1.1 | Real-time analytics via WebSocket streaming |
| ๐ Planned | v1.2 | Multi-user collaborative workspaces with role-based access |
| ๐ Planned | v1.3 | ERP connectors โ SAP, Oracle NetSuite, Microsoft Dynamics |
| ๐ Planned | v1.4 | CRM connectors โ Salesforce, HubSpot, Pipedrive |
| ๐ Planned | v2.0 | Explainable AI with full causal reasoning chains |
| ๐ Planned | v2.1 | Enterprise SSO โ Google Workspace, Okta, Azure AD |
| ๐ Planned | v2.2 | One-click PowerPoint and PDF boardroom export |
| ๐ Vision | v3.0 | Google Vertex AI integration with custom fine-tuned models |
Pravalika Palle
Lead Engineer ยท Product Designer ยท AI Architect
MCA Graduate ยท AI & Data Enthusiast ยท Builder of SynapseIQ
This project is licensed under the MIT License โ see the LICENSE file for details.
Copyright ยฉ 2026 Pravalika Palle
โก SynapseIQ
Where Data Becomes Decisions
AI Operating System for Executive Decision Making
Powered by Google Gemini ยท Built for Google GenAI Academy Hackathon 2026
If you found this project interesting, please consider giving it a โญ on GitHub.






