Skip to content

Nexus AI

Nexus AI is a production-ready, multi-agent orchestration platform built on Google Cloud Run that automates and audits corporate claims with absolute trust.

Code Video Cerebral Valley

Core trick

Gemini Flash extracts invoice fields, parallel agents check policy limits and duplicates, and a circuit breaker pauses claims for supervisor approval.

Tags

financeagents

Description

Nexus AI is a production-ready, multi-agent orchestration platform built on Google Cloud Run that automates and audits corporate claims with absolute trust. The Engine: Invoices are ingested using Gemini Flash for high-fidelity OCR, instantly launching a parallel agent grid (Provider Registry, Policy limits in ₹, and Pattern fraud scans). If overruns or duplicate anomalies are flagged, a Human Gate Circuit Breaker freezes execution to route the claim for supervisor approval. The Auditor: An independent Gemma compliance node (gemma2-27b-it) verifies the decision logic, generating targeted stakeholder summaries and a synthesized Text-to-Speech (TTS) voice briefing so supervisors can listen to case alerts instantly. Fully integrated with real-time SSE streaming, Nexus is the first platform to decouple core adjudication from auditing—combining high-speed parallel intelligence with bulletproof compliance and human-in-the-loop control.
Kiruthick7/nexusAI★ 00 forksPythonCreated Jul 11, 2026 · last push Jul 11, 2026

README excerpt

# Nexus AI Operations Platform

> **The Multi-Agent Orchestration & Decision Intelligence Engine for Autonomous Claim Adjudication.**

Welcome to the **Nexus AI Operations Platform**, a state-of-the-art enterprise-grade multi-agent platform designed to automate, audit, and optimize complex claims processing. Powered by **Google Gemini** models and independent **Gemma** intelligence layers, Nexus AI translates raw invoice documents into deterministic adjudications through a coordinated, parallelized, and highly explainable multi-agent system.

---

## 🏛️ System Architecture

The Nexus AI platform is engineered using a **Stateless Event-Driven Decoupled Architecture**. Claims undergo multi-stage topological orchestration while publishing real-time telemetry back to the client via Server-Sent Events (SSE).

```mermaid
graph TD
    Client["Client Dashboard (React/Next.js)"]
    API["FastAPI Gateway (Uvicorn)"]
    DB[("Mission Database (In-Memory / PostgreSQL)")]
    Bus["Central Event Bus (SSE Publisher)"]
    
    Intake["Intake Agent (Document Parser)"]
    Planner["Planner Agent (Orchestration Graph)"]
    
    subgraph Specialists ["Parallel Specialist Pool"]
        Provider["Provider Agent (Ledger Check)"]
        Policy["Policy Agent (Limit Enforcer)"]
        Pattern["Pattern Agent (Risk Scanner)"]
    end
    
    Arbiter["Arbiter Decision Engine"]
    Resolver["Evidence Resolver (Conflict & Weights)"]
    Gates["Tool Gates (Circuit Breakers)"]
    
    Gemma["Gemma Int

Team