IT Glossary
What is Serverless Computing?
Code with no servers to manage: the cloud platform runs your functions on demand, scales automatically and bills by the millisecond.
The name is a small lie told for convenience: the servers still exist, but they belong to the cloud platform and nobody on your payroll ever touches them. Developers write only the functions — "when an order arrives, generate the invoice"; "when a file is uploaded, resize the image" — and the platform executes each one on demand, scales from zero to thousands of parallel runs automatically, and charges for the milliseconds of compute actually consumed. That pricing model is why finance teams like it: a workload used sporadically costs close to nothing while idle, making serverless ideal for event-driven processing, APIs with irregular traffic and scheduled automations. The trade-offs are known and manageable: at heavy constant load, per-execution pricing can overtake a plain server; the first call after a quiet period adds a short "cold start" delay; and functions written around one platform's conventions take effort to move elsewhere. Two questions settle most vendor conversations: what does this cost at our volumes — and at ten times our volumes — and what would leaving the platform involve?
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247
Why it matters for your business
Almost no fixed operating costs
You pay only for executions consumed — sporadic and seasonal workloads cost pocket change per month, not the price of an idle server.
Automatic scaling from zero to peak
Thousands of parallel executions during a campaign, none overnight — no capacity planning and no 2 a.m. interventions.
Teams ship features, not maintenance
No patching, sizing or infrastructure worries — development hours go entirely into your business logic.
Frequently asked questions
Is serverless always the cheaper option?
No — it is proportional to usage: unbeatable for sporadic or variable traffic, but at heavy constant load the arithmetic flips and classic servers or containers become more economical again. Ask for a cost estimate on your real traffic profile, plus the growth scenario, with the break-even point identified before you sign.
What are cold starts and should we worry about them?
A function unused for a while goes dormant, and the first execution afterward pays a wake-up penalty of tens of milliseconds to a few seconds depending on the platform. Irrelevant for background jobs and most business APIs; for latency-sensitive interactions it is mitigated with kept-warm instances or a different model for that one component.
Does serverless lock us into one cloud provider?
Partially, and manageably: disciplined business logic separated from platform conventions moves reasonably well; the real ties are the surrounding services — queues, databases, authentication from the same vendor. Keep proprietary services only where they add clear value and verify data export periodically; lock-in is not eliminated, it is kept at a consciously accepted level.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247