Pulsyon is a backend systems engineering project for telemetry ingestion and incident analysis. It explores relational schema design, SQL query performance, REST API development, asynchronous processing, and system evolution through a phased V1 → V4 architecture.
Built as part of the Google Software Engineering Program (BASTA G-SWEP), with mentorship from a Google software engineer.
Pulsyon simulates the backend of a developer-facing observability tool. It is designed to answer:
- Are my services healthy?
- What events are happening in my system?
- How does performance change under load?
- (Later) What incidents are currently active and why?
The project is intentionally built in stages. Each version is independently runnable and introduces a new layer of system complexity.
Current Implementation: V1: Foundation
V1 focuses on:
- PostgreSQL schema design and migrations
- SQL query development and validation
- REST API with basic CRUD endpoints
- Seed data generation
- Query performance analysis using
EXPLAIN ANALYZE
Not included yet:
- Redis / async processing
- background workers
- incident detection
- deployment / containerization
- relational schema (PostgreSQL)
- core tables and indexing
- CRUD + query endpoints
- seed data + analytics queries
- performance baseline analysis
- Redis-based ingestion
- background worker(s)
- dead-letter handling
- authentication + multi-tenancy
- aggregation pipelines
- anomaly detection
- incident grouping logic
- analytics endpoints
- caching + load testing
- lightweight dashboard
- Docker / containerization
- CI/CD with GitHub Actions
- cloud deployment (GCP Cloud Run)
- system monitoring and reliability improvements
- Backend: Node.js, Express
- Database: PostgreSQL
- Queue / Cache: Redis (introduced in V2/V3)
- Performance Testing: k6
- Deployment: Docker, GitHub Actions, GCP Cloud Run
- Documentation: ADRs, architecture diagrams, performance reports
Pulsyon is built around a structured telemetry and analytics schema:
users: authentication and rolesservices: registered systems emitting eventsapi_logs: raw telemetry events (partitioned)service_health: computed service metricsmetrics_hourly: aggregated performance dataincidents: grouped anomaly events (V3)alerts: individual anomaly triggers (V3)failed_events: dead-letter storage (V2)
In V1, the focus is validating schema design and query performance before introducing asynchronous infrastructure.
- Node.js
- PostgreSQL
git clone <repo-url>
cd pulsyon
npm install
npm run migrate
npm run seed
npm run dev