Skip to content

Platform Architecture Overview

Platform Architecture Overview

The SDGL-SaaS Platform uses a six-layer architecture that forms an integrated, modular system designed for scale, reliability, and continuous learning.

Six-Layer Architecture

┌─────────────────────────────────────────────────────────────┐
│ LAYER 1: SOURCE LAYER │
│ (Data Collection & Integration) │
└─────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────┐
│ LAYER 2: INTEGRATION LAYER │
│ (Data Normalization & Consolidation) │
└─────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────┐
│ LAYER 3: PROCESSING LAYER │
│ (Analytics, ML, Computations) │
└─────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────┐
│ LAYER 4: STORAGE LAYER │
│ (Persistent Data & Audit Trail) │
└─────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────┐
│ LAYER 5: ANALYTICS LAYER │
│ (Aggregation, Reporting, Insights) │
└─────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────┐
│ LAYER 6: VISUALIZATION LAYER │
│ (Dashboard, Reports, User Interface) │
└─────────────────────────────────────────────────────────────┘

Each layer builds upon the previous, enabling:

  • Separation of Concerns: Each layer has specific responsibilities
  • Scalability: Can independently scale any layer
  • Reliability: Failures in one layer don’t cascade
  • Innovation: New algorithms/interfaces don’t require full rewrite

Layer 1: Source Layer →

Data collection from diverse sources (surveys, IoT, systems, databases, APIs)

Layer 2: Integration Layer →

Normalize and consolidate data into unified schema

Layer 3: Processing Layer →

Run analytics, machine learning, and complex computations

Layer 4: Storage Layer →

Persistent database with full audit trail and compliance

Layer 5: Analytics Layer →

Aggregate results, identify patterns, generate insights

Layer 6: Visualization Layer →

Present insights via dashboards, reports, and user interfaces


Modular Microservices Architecture

The platform is built on microservices that operate independently:

Core Services

  • Authentication Service: User identity, permissions, SSO
  • Assessment Service: ESGETC scoring, materiality calculations
  • Planning Service (PDCA): Goal setting, progress tracking, learning
  • Consortium Service: Multi-stakeholder partnerships, Delphi process
  • Entity Service: Discovery, classification, relationship mapping
  • Analytics Service: KPI calculations, reporting, insights
  • Integration Service: External systems, APIs, webhooks

Supporting Services

  • Notification Service: Alerts, emails, in-app messages
  • Reporting Service: PDF/Excel exports, customizable templates
  • Search Service: Full-text and semantic search
  • File Service: Document upload, storage, OCR
  • Audit Service: Compliance logging, version control
  • Cache Service: Performance optimization, real-time updates

Each microservice:

  • Owns its database (no shared data stores)
  • Uses REST APIs for inter-service communication
  • Scales independently based on demand
  • Can be deployed/updated independently
  • Has explicit versioning and backward compatibility

Data Flow & Processing Pipelines

Real-Time Data Pipeline

IoT Sensors → Kafka → Real-time Processor → Cache/Dashboard
(production equipment) (message queue) (anomaly detection)

Batch Processing Pipeline

Surveys → Data Lake → Python/Spark Jobs → Analytics DB → Reports
(user input) (staging) (transformations) (aggregated)

Machine Learning Pipeline

Tagged Training Data → ML Models → Prediction Serving → Dashboard
(historical assessments) (NLP, clustering, forecasting) (recommendations)

Entity Discovery Pipeline

Multiple Sources → NLP Extraction → Deduplication → Verification → Platform
(Web, DBs, APIs) (entity recognition) (pairwise matching) (human review)

Learn more about data pipelines →


Security, Privacy, and Compliance Architecture

Security Layers

  • Network: TLS/SSL encryption, VPN, firewall rules
  • Application: Input validation, SQL injection prevention, XSS protection
  • Data: Row-level security, field-level encryption, API keys
  • Identity: Multi-factor authentication, role-based access control
  • Audit: All actions logged with user/timestamp/details

Privacy Controls

  • Data Minimization: Collect only what’s necessary
  • Anonymization: Benchmark data stripped of identifiers
  • Retention: Automatic deletion after policy period
  • Portability: Export personal data in standard formats
  • Right to be Forgotten: Comply with GDPR and regional laws

Compliance Standards

  • SOC2 Type II: Security and availability audit
  • GDPR: EU data protection compliance
  • HIPAA: Health data confidentiality (where applicable)
  • SDG Frameworks: Alignment with UN standards
  • ESG Standards: GRI, SASB, TCFD reporting

Learn more about security and compliance →


Core Technology Stack

Frontend

  • React 18: Modern UI framework with hooks
  • TypeScript: Type-safe JavaScript
  • Tailwind CSS: Utility-first styling
  • D3.js / Three.js: Data visualization and 3D rendering
  • Redux: State management for complex UIs

Backend

  • Node.js: JavaScript runtime for servers
  • TypeScript: Type-safe backend code
  • Wasp: Full-stack framework (TypeScript + Node + React)
  • Express/Fastify: HTTP servers
  • GraphQL: Flexible API queries (optional)

Data & AI

  • PostgreSQL: Primary relational database
  • Redis: Caching and real-time updates
  • Python/Pandas: Data processing and analytics
  • Spark: Big data processing at scale
  • TensorFlow/PyTorch: Machine learning models
  • spaCy/Hugging Face: NLP for entity classification

Infrastructure

  • Docker: Container deployment
  • Kubernetes: Container orchestration (production)
  • GitHub Actions: CI/CD pipelines
  • AWS/GCP/Azure: Cloud platforms
  • Terraform: Infrastructure as Code

Scalability & Performance

Handles Growth

  • Users: From 100s to 100,000+ concurrent users
  • Data: From millions to billions of data points
  • Assessments: From dozens to 100,000+ organizations
  • Partnerships: From small consortiums to country-wide networks
  • Real-time: 1,000+ events per second through IoT pipeline

Performance Optimization

  • Asynchronous Jobs: Long-running tasks don’t block UI
  • Caching Layers: Frequently accessed data pre-computed
  • Database Optimization: Indexes, partitioning, query optimization
  • CDN: Global distribution of static assets
  • API Rate Limiting: Prevent abuse, ensure fairness

Monitoring & Observability

  • Prometheus/Grafana: Metrics collection and visualization
  • ELK Stack: Centralized logging (Elasticsearch)
  • Distributed Tracing: Understand performance bottlenecks
  • Alerting: Proactive notification of issues
  • Incident Response: Runbooks for common issues

Integration Points

External Data Sources

  • World Bank Indicators
  • UNDP SDG databases
  • ESG rating agencies
  • Country statistical bureaus
  • OpenStreetMap
  • Wikidata
  • Custom REST APIs

Business Systems

  • ERP: SAP, Oracle, NetSuite integration
  • CRM: Salesforce, HubSpot
  • HR: Workday, BambooHR
  • Accounting: QuickBooks, FreshBooks
  • Project Management: Asana, Jira, Monday

Communication

  • Email (Gmail, Office 365)
  • SMS/WhatsApp
  • Slack webhooks
  • Teams integration
  • Custom webhooks

Payment

  • Stripe for subscription billing
  • Wire transfers for bulk contracts
  • Government purchasing card integration
  • Cryptocurrency (future roadmap)

Next Steps