FranchiseRamp Campaign Automation Platform
Enterprise-grade distributed microservices ecosystem for multi-platform ad campaign automation and analytics.
An end-to-end campaign automation and data synchronization ecosystem designed for franchisors. It unifies identity management, automated ad insights ingestion from Facebook and TikTok APIs, high-throughput lead processing, and predictive suggestion generation.
Technology stack
Problem
Franchisors faced massive manual operational overhead in managing cross-channel marketing campaigns, unifying lead data across hundreds of locations, and deriving actionable ad performance metrics in real time.
Solution
Built an event-driven microservices platform utilizing Celery, RabbitMQ, and Elasticsearch to automate daily ad data synchronization, compute multi-period CPL/ROI performance metrics, and generate real-time operational suggestions for account managers.
Technical contribution
Designed and documented the distributed backend microservices architecture, authored production-grade Elasticsearch indices and aggregation pipelines, implemented RBAC with JWT token lifecycle management, and built asynchronous task execution pipelines.
Architecture
Event-driven microservices architecture featuring an Auth IAM service (Django/PostgreSQL/RabbitMQ), a core Campaign Automation orchestrator (FCA), a dedicated Data Sync Engine (Python/Elasticsearch), and a centralized Celery scheduler with Flower monitoring.
Performance considerations
Optimized high-volume write operations using Elasticsearch async bulk ingestion, and implemented year-based dynamic index sharding to maintain low-latency query aggregations across multi-year dataset histories.
Security considerations
Implemented strict stateless JWT authentication with automated refresh-token blacklisting, fine-grained RBAC for functional scopes (ad management, reporting), and secure service-to-service communication via internal API keys.
Deployment model
Containerized multi-service setup managed with Docker Compose locally and Kubernetes (Helm) in production, powered by GitLab CI/CD pipelines with Flake8 linting and automated coverage reporting.
Outcome
Automated 90%+ of manual ad data sync workflows, enabled instant cross-platform performance visibility across thousands of franchise locations, and provided account strategists with daily automated budget and optimization suggestions.