Deployment + Development for RCS & Deployment for WhatsApp

Deployment + Development for RCS & Deployment for WhatsApp
1. SOW for deployment for RCS & WhatsApp

This defines the requirements and scope of work for setting up, deploying, and scaling the
CPaaS Dashboard. The system is a microservices architecture with PostgreSQL as the main
database. Currently, the WhatsApp and RCS services have been built and integrated.

Objectives & Outcome

The objective is to engage a DevOps to configure infrastructure, databases, queues,
monitoring, and deployment pipelines to ensure the CpaaS dashboard is production-ready,
scalable.

Post award of project you will be shared a Github Repo weekly or daily update to be pushed there
only. The scope includes backend API development, database schema design, & webhook
hooks. The service must be built as a self-contained microservice with a well-defined REST API
to ensure seamless, scalable, and repeatable integration into the main platform.

Performance and reliability targets define the acceptance criteria for this engagement:

• Sustain 500–1,000 API send requests per second under steady load
• Gracefully absorb burst traffic up to 3,000 TPS via queue buffering without message loss
• Maintain callback processing latency below 2 seconds end-to-end
• Enable deployment rollback within 2 minutes with service continuity
• Provide automated retry handling and dead-letter workflows for failed messaging jobs
• Deliver queue, database, and API observability sufficient for rapid fault diagnosis

Framework of the project

Frontend
• Framework: React.js
• Language: TypeScript

Backend
• Runtime: Node.js
• Language: TypeScript
• Framework: Express.js
• API Style: REST
• ORM layer: Drizzle ORM
• Job Queue: BullMQ + Redis
• In-memory datastore: DragonflyDB (Redis-compatible)
• Authentication: BetterAuth
• Package Manager: pnpm

Database: PostgreSQL (DigitalOcean managed database)
Hosting Platform: Digital Ocean

Scope of Project
Infrastructure Setup

Provision cloud infrastructure on DigitalOcean.

Create separate environments for Development and Production.

Deploy microservices as Docker containers for RCS & WhatsApp.

Configure reverse proxy (NGINX/Caddy) with SSL certificates.

Database Setup

Set up DigitalOcean Managed PostgreSQL cluster.

Create separate databases or schemas for each microservice (RCS, WhatsApp).

Enable PgBouncer connection pooling.

Configure automated backups, replication, and monitoring.

Caching & Queues

Install and configure Dragonfly (Redis alternative) for caching, sessions, and queues.

Create topics for RCS, WhatsApp.

Implement monitoring for queue depth, lag, and retries.

Deployment Pipeline

Set up CI/CD pipelines using GitHub Actions.

Enable zero-downtime deployments (rolling updates / blue-green).

Define separate workflows for Development and Production.

Ensure automatic deployment to staging on dev branch and production on main branch.

Monitoring & Logging

Set up centralized logging (Grafana Loki or ELK Stack).

Implement monitoring dashboards (Prometheus + Grafana).

Configure alerts for queue failures, DB overload, API latency, and service downtime.

Real-time notification integration (Email).

Scaling & Performance

Design for horizontal scaling of API workers and queue consumers.

Enable TPS control and rate limiting.

Optimize message pipelines for millions of requests/day.

Create dead-letter queues for failed jobs.

Security & Compliance

Enforce HTTPS-only communication.

Apply firewall rules to restrict database access.

Implement data privacy standards (IP masking, anonymized logs).

2. Sow for development of RCS

We want to add full RCS messaging capabilities as a separate microservice that supports
both Public API & UI, allowing us to route messages through different providers. The
objective of this project is to extend our platform by building a dedicated, production-grade
messaging microservice responsible for handling WhatsApp and RCS message delivery
through APIs. This service must expose scalable public APIs.

The public APIs should be architected for high throughput, modular provider routing, and
future extensibility, enabling us to support millions of campaign messages without
requiring structural.

a. API & Webhooks for Clients

-Public API endpoints for sending campaigns, managing templates, and fetching reports.

Webhook support for:
-Delivery receipts (DLRs)
-Incoming messages/user replies

-Provide an OpenAPI spec or tRPC schema for client integrations.

b. Integration Layer

-Built-in connectors for MoEngage, CleverTap, and other CRMs.
-Ability to easily add more integrations later.