The words, without the jargon.

Authentication

Establishing who a user is — the login step. Working authentication tells you nothing about whether authorization is also enforced.See also: Authorization

Authorization

Deciding what an already-identified user is allowed to do. Distinct from authentication, which decides who they are. Most data exposure in AI-built applications is an authorization failure, not an authentication one.See also: Authentication, Row-level security

CI/CD

Continuous integration and continuous delivery — automation that tests every change and deploys it the same way every time. Its practical value is that releases become boring and reversible.

Database index

A lookup structure that lets a database find matching rows without examining every row. Missing indexes are the single most common cause of an application that gets slower as it gets more successful.

Idempotency

The property that performing the same operation twice has the same effect as performing it once. Essential for payments and webhooks, where retries and duplicate deliveries are normal rather than exceptional.

Insecure direct object reference

A flaw where changing an identifier in a request — a number in a URL, an ID in a payload — returns data belonging to someone else, because the server never checked that the caller owns the record.

N+1 query

Fetching a list with one query, then issuing another query per item in that list. Imperceptible with ten records and crippling with ten thousand, which is why it survives testing and fails in production.

Production-readiness score

IndiaNIC's assessment of an application against 49 checks in 8 weighted categories, expressed out of 100. It exists to turn "is it ready?" into a specific, ranked list of what is missing.

Production-ready

Software that can be operated by people other than its author, under real traffic, with real data, and recovered when something goes wrong. It is a property of the whole system — security, testing, deployment, monitoring — not of the features.

Prototype

A build whose purpose is to answer a question — will this work, will anyone want it. A prototype is successful when it has answered the question, which is a different bar from being safe to operate.

Rate limiting

Capping how often a caller can hit an endpoint. Protects against both abuse and the accidental loop that turns a bug into an invoice.

Rollback

Returning to the previously working version after a bad release. A rollback path only counts if it has been tested — an untested one tends to be discovered as unusable at the worst moment.

Row-level security

Access rules applied by the database itself, per record, so a query returns only the rows the caller is entitled to. Because it sits below the application, it holds even when a request bypasses the interface.See also: Authorization

Staging environment

A complete copy of production, with non-production data, where changes are exercised before real users see them. Without one, production is the staging environment.

Technical debt

The accumulated cost of decisions that were faster at the time. Not inherently bad — debt taken deliberately to reach a milestone is a legitimate trade. It becomes a problem when nobody recorded that it was taken.

Vendor lock-in

Dependence on one platform's proprietary features to the point where leaving means rebuilding. The cost is invisible while you are happy with the platform.

Vibe coding

Building software by describing what you want in natural language and letting an AI tool generate the implementation, judging the result by whether it appears to work rather than by reading the code. The term describes a way of working, not a level of quality — the output can be excellent or unshippable.

Webhook

A message sent from one system to another when something happens — a payment succeeding, a subscription lapsing. Payment integrations that never handle webhooks appear to work until the first event that is not an immediate purchase.See also: Idempotency

These terms all come up in the production-readiness framework. If you want to see how they apply to the specific tool that built your application, start with the platform guides — or read what production engineering actually involves.