🌐 Crisis Topography: Command Center

Reframing Global Humanitarian Aid as a Quantitative Capital Allocation Challenge.

Humanitarian aid allocation is currently reactive, driven by media sentiment rather than systemic need. Crisis Topography is a 3D spatial intelligence system designed to identify, analyze, and predict humanitarian crises through a visceral, agentic interface.

🤖 The Magnum Opus: Meet Pablo

Pablo is not just an AI; he is the sovereign orchestrator of the Crisis Topography Command Center.

Commanded solely through natural language voice interaction, Pablo serves as the bridge between human intuition and machine intelligence. He doesn't just answer questions—he drives the entire system.

• Vocal Autonomy: Utilizing the ElevenLabs Voice Isolator and high-fidelity WebRTC streams, Pablo hears through the noise of any crisis environment.
• Intuitive Navigation: Tell Pablo, "Take me to the border of Sudan," and he will physically fly the 3D globe to the precise location, adjusting altitudes and atmospheric glow to highlight semantic anomalies.
• Agentic Synthesis: Pablo cross-references Databricks analytical clusters with Actian vector embeddings to deliver real-time, personalized intelligence briefings.
• Command Feedback Loop: When a user triggers a complex query, Pablo orchestrates the ensemble of Gemini, Llama, and Kafka Streams, reporting back not just with data, but with a strategy.

"Pablo turns a cold data dashboard into a living, breathing tactical partner."

---

🏗️ The Distributed Agentic Pipeline (The "Engine")

Our architecture is a multi-stage, event-driven ecosystem where voice, vision, and vector intelligence converge. We don't just process data; we orchestrate a resonance chamber of AI models and high-throughput data streams.

graph TD
    User((User/Crisis Manager)) -- Voice/WebRTC --> EL1[ElevenLabs Isolator]
    EL1 -- Denoiser/Isolator --> EL2[ElevenLabs Voice Ingress]
    EL2 -- Stream --> Gem1[Gemini 1.5 Pro]
    Gem1 -- Function Caller --> KS1[Kafka Stream: Command Bus]
    KS1 -- Dispatch --> Client[Client Side Command Trigger]
    Client -- Logic --> Ensemble{Model Ensemble}
    Ensemble -- Analytical --> Genie[Databricks Genie/LLama]
    Ensemble -- Retrieval --> Actian[Actian Vector DB]
    Genie -- Results --> KS2[Kafka Stream: Result Aggregator]
    Actian -- Vectors --> KS2
    KS2 -- Sync --> Client
    KS2 -- Callback --> Gem2[Gemini 1.5 Flash: Narrative Synthesis]
    Gem2 -- SSML/Text --> EL3[ElevenLabs Voice Egress]
    EL3 -- Audio --> User

Loading

🔄 The Seven-Phase Resonance Loop

1. Acoustic Isolation & Ingress: High-fidelity voice input is captured via WebRTC and passed through the ElevenLabs Voice Isolator. Noise is stripped, and intent is clarified before hitting our primary LLM orchestrator.
2. Agentic Dispatch (Gemini 1.5 Pro): Gemini acts as the "Central Nervous System," performing real-time function calling to decompose natural language into executable system commands.
3. Command Bus (Kafka Streams): Commands are published to a high-concurrency Kafka Stream. This ensures sub-millisecond dispatching to client-side triggers and backend analytical engines.
4. Multi-Model Ensemble Execution:
   • Analytical Reasoning: Complex SQL queries are generated and executed against Databricks Genie Spaces for deep structured data mining (Llama/Gemini).
   • Semantic Retrieval: High-dimensional embeddings of 8,000+ humanitarian projects are searched within the Actian Vector DB using actiancortex.
5. Asynchronous Aggregation: Results from Databricks and Actian are fed back into a secondary Kafka Stream, serving as a global state synchronizer for both the 3D frontend and the narrative generator.
6. Narrative Synthesis (Gemini 1.5 Flash): The raw data—anomaly scores, ROI deltas, and crisis metrics—is synthesized into a human-readable "Intelligence Briefing" by Gemini 1.5 Flash.
7. Voice Egress & Haptic Feedback: The briefing is streamed back via ElevenLabs, while the 3D globe performs Topographic Extrusion—physically deforming the planet's geometry to represent the "Height of Suffering" (unmet funding gaps).

---

🛠️ Technical Deep Dive

The Intelligence Core

• Actian Vector DB: Our "Long-Term Memory." We vectorize global project portfolios to perform semantic benchmarking, revealing where aid "yield" is highest.
• Databricks Lakehouse: The bedrock of our structured intelligence, providing real-time ingestion from UN OCHA/HDX HAPI v2.
• Vector Architecture: Utilizing sentence-transformers/all-mpnet-base-v2 for dense vector representations of geopolitical narratives.

The Visual Layer

• Next.js 14 & Three.js/WebGL: A custom-engineered 3D environment where data is not just plotted, but physically manifested as interactive geometry.
• Topographic Mismatch Engine: A proprietary algorithm that calculates the Severity-to-Funding Gap Ratio, projecting these as 3D extrusions on the globe.

The Communication Stack

• ElevenLabs WebRTC Agent: Low-latency, bidirectional voice interaction that allows users to "talk to the planet."
• FastAPI Backend: Built for extreme concurrency, managing the bridge between our event streams and the generative models.

---

☁️ Infrastructure & Deployment

To handle the high-concurrency demands of a real-time global intelligence system, we utilize a distributed cloud architecture.

• Vultr High-Compute Nodes: Our backend and Actian Vector DB are orchestrated on Vultr's optimized compute instances. This provides the dedicated CPU performance required for sub-100ms vector similarity searches and rapid geospatial calculations.
• Actian on Vultr: Unlike shared cloud databases, our self-hosted Actian instance on Vultr allows for granular performance tuning, ensuring that the heavy query load from the 3D globe does not bottleneck our agentic response times.
• Databricks Lakehouse Integration: Seamlessly bridging the gap between high-scale data engineering in Databricks and high-speed execution on Vultr.

---

🎯 Track Pursuit Strategy

• 🏆 Best Overall: A seamless fusion of WebGL visuals, voice-first agentic UX, and enterprise-grade data architecture.
• 💰 Finance: Revolutionizing humanitarian aid by treating global crises as a portfolio optimization problem with ROI metrics.
• 🛡️ SafetyKit & Actian: Leveraging vector anomalies to predict systemic risks (mass migration, famine) before they escalate into catastrophes.
• 🎨 Pure Imagination: A world where you don't read a dashboard—you navigate a living, breathing topography of data.

---

🚀 Getting Started

# Clone the repository
git clone https://github.com/sairamen/Hacklytics-GoldenByte

# Install Backend Dependencies
cd backend
pip install -r requirements.txt

# Install Frontend Dependencies
cd ../frontend
npm install

# Start the Command Center
npm run dev

---

"The map is not the territory. But a predictive, intelligent map can save the territory before it burns."