# Why we reach for Rust on operational systems

Canonical: https://collabwire.io/blog/rust-for-operational-systems

## Summary

Field note on when Collabwire uses Rust for operational systems — throughput, long-running workers, IoT and ledger logic — and when it does not.

## Excerpt

Rust is not a religion. It is a tool we pick when concurrency, throughput and long uptime matter more than how fast the first prototype ships.

## Tags

rust, engineering, architecture

## Content

We do not start every project in Rust. Most operational software is fine in TypeScript, Go or Python. We reach for Rust when the cost of a crash, a memory leak or a slow worker is higher than the cost of moving slower in week one.


## Where Rust earns its keep


- Event pipelines that need to absorb traffic spikes without backpressure collapse.
- Long-running workers where a 0.3% leak per hour turns into a Sunday outage.
- IoT and ingestion layers where the cost of a single dropped message is real money.
- Financial and ledger logic where overflow, rounding and concurrency bugs are not negotiable.


## Where Rust is the wrong answer


Internal admin tools, content sites, low-traffic CRUD. The team will move twice as slow for no real benefit. Pick the runtime that matches the failure mode you actually have, not the one that looks good in a conference talk.


> Rust pays you back in the third year, not the third week.