The Messaging Backend Behind Millions of Monthly Texts
A US consumer app gives its users a second phone number for texting and calling. NerdDevs builds and operates the backend that moves those messages — millions of them, every month, for years.
- Client
- US app studio — name withheld by agreement
- Engagement
- Multi-year build & operate partnership
- Scale
- Millions of texts delivered monthly
Overview
Our client — a US app studio whose name we withhold by agreement — runs a consumer messaging product: a second phone number that lives in an app, used for texting and calling when people don’t want to hand out their real one. Behind that simple promise sits an unforgiving engineering problem — messages must arrive quickly, reliably, and at a unit cost that leaves room for a viable business.
NerdDevs is the long-term engineering partner behind that backend. We designed the message pipeline, run the AWS infrastructure it lives on, and carry the pager for it — a build-and-operate engagement that has lasted years and now moves millions of texts a month.
The Challenge
Carrier-grade expectations
A text that arrives late — or not at all — is a broken product. Users judge a second phone number against their real one, so deliverability is the feature.
Bursty, always-on traffic
Messaging load spikes with campaigns, holidays and time zones while the platform runs 24/7 for a US audience — there is no maintenance window that suits everyone.
Abuse and compliance pressure
A texting product attracts spam and abuse. Keeping number reputation clean and staying inside carrier compliance rules is continuous operational work, not a one-time setup.
Cost per message at scale
At millions of texts a month, small inefficiencies in routing, retries and storage turn into real money. Unit economics have to be engineered, not hoped for.
What We Build & Operate
Message pipeline with queues & retries
Queued delivery with idempotent retries and dead-letter handling, so transient provider failures never silently drop a user’s text.
Load-balanced AWS delivery tier
Auto-scaling application servers behind load balancers with health checks, sized against real traffic patterns and burst-tested.
Abuse & reputation controls
Rate limiting, spam heuristics and number-reputation hygiene baked into the pipeline to protect deliverability for legitimate users.
Observability & on-call operations
Structured logs, metrics, alerting and runbooks — we operate what we build, and have for years as the client’s long-term engineering partner.
Impact
- Millions of texts delivered every month for a US consumer audience.
- A multi-year build-and-operate partnership — the same team, years later, still shipping.
- 99.9% uptime as the operating bar, consistent with the platforms we run across our portfolio.
- Messaging unit economics kept predictable at scale through routing, retry and storage discipline.
Why no client name? Confidentiality is part of the engagement — the same discretion we’ll extend to your product.
Technology Stack (infrastructure view)
- Cloud
- AWS
- Backend
- Node.js
- Traffic & scaling
- Load balancers with health checks, auto-scaling groups
- Pipeline
- Redis-backed queues, idempotent retries, dead-letter handling
- Data
- MongoDB, Redis
- Operations
- CI/CD, structured logging, metrics & alerting, on-call runbooks
What would millions of monthly requests do to your infrastructure?
We’ll tell you before your users find out. A 5-day Cloud Cost & Security Audit reviews your setup with the same discipline we apply here — and if we don’t find savings worth 3× the fee, it’s free.
Book Your Cloud Audit →