Intellora Tech
ProductsProjectsToolsBlogAboutPayments Price estimator Book a call

Glossary

Data engineering, in plain English.

32 terms we use on this site and in real project conversations, explained without the jargon that usually surrounds them.

ETL

Data is pulled from a source, reshaped to fit the destination, then loaded in. The traditional order for moving data into a warehouse.

Database Engineering

ELT

Raw data lands first, and the transformation happens afterwards, inside the warehouse itself. The more common pattern on modern cloud platforms.

AWS Cloud Architecture

Data lake

A single store for raw data in its native format, structured and unstructured alike, before anyone decides exactly how it will be used.

AWS Cloud Architecture

Data warehouse

A store built around how the business asks questions: modelled, structured, and optimised for reporting rather than for capturing every raw detail.

AWS Cloud Architecture

Data lakehouse

A lake and a warehouse combined: the flexibility of raw storage with the structure and performance guarantees of a warehouse on top.

AWS Cloud Architecture

Data pipeline

The automated sequence that moves data from source to destination: extract, transform, validate, load, on a schedule or in response to an event.

Database Engineering

Data lineage

The record of where a piece of data came from and every transformation it passed through on the way to a report. What an auditor asks for first.

Data Governance

Data governance

Who owns which data, who can change its definition, and what evidence proves the rules were actually followed, not just written down.

Data Governance

Data catalog

A searchable inventory of what data exists, what it means, who owns it and how fresh it is, so people stop asking in Slack.

Data Governance

Data quality

Whether data is accurate, complete, consistent and timely enough for the decision being made with it. Usually enforced with automated checks.

Data Governance

Master data management

Keeping one agreed, trustworthy version of a core entity, a customer, a product, a supplier, instead of five slightly different copies across systems.

Data Governance

Semantic layer

The place a metric is defined once, so every dashboard and report reads the same definition instead of recalculating it slightly differently.

Analytics & BI

Data mart

A smaller, focused slice of the warehouse built for one department or subject area, so they aren't querying the entire estate for one answer.

Analytics & BI

Schema

The structure of a dataset: table names, columns, data types and how they relate. What a schema change breaks when nobody is warned first.

Database Engineering

Schema drift

A source system silently adds, removes or retypes a column, and anything downstream that assumed the old shape quietly breaks or miscounts.

Data Governance

Change data capture

Streaming only the rows that changed, rather than re-reading an entire table each time. How near-real-time pipelines stay fast at scale.

AWS Cloud Architecture

Orchestration

The scheduler that decides what runs, in what order, and what happens on failure, Airflow and similar tools, rather than a chain of cron jobs.

AWS Cloud Architecture

Data observability

Monitoring not just whether a job ran, but whether the data it produced is actually correct, fresh and complete.

Data Governance

High availability

A system designed to keep running, or fail over automatically, when a component breaks, instead of going down until someone notices.

Database Engineering

Disaster recovery

The tested plan, and the tested backups, for getting back up after a serious failure. Untested backups are not a recovery plan.

Database Engineering

Replication

Keeping a live copy of a database in sync with the original, for failover, read scaling, or feeding another system in near real time.

Database Engineering

Landing zone

The governed, pre-built account and network structure a cloud workload moves into, so every new environment starts from the same secure baseline.

AWS Cloud Architecture

FinOps

The discipline of treating cloud cost as an engineering problem: tagging, right-sizing and architecture decisions, not just a bill to complain about.

AWS Cloud Architecture

Infrastructure as code

Infrastructure defined in version-controlled files (Terraform, CloudFormation) instead of clicked together by hand, so it can be reviewed and repeated.

AWS Cloud Architecture

Well-Architected Framework

AWS's own structured review across operational excellence, security, reliability, performance, cost and sustainability.

AWS Cloud Architecture

Feature store

A central place that stores the inputs machine learning models train on, so the same feature is defined identically for training and for live predictions.

AI & Machine Learning

MLOps

The deployment pipelines, monitoring and retraining schedules that keep a model working in production, the operational half of machine learning.

AI & Machine Learning

Model drift

A model's accuracy quietly degrading over time as the real world stops matching the data it was trained on, until nobody trusts its answers.

AI & Machine Learning

RAG

Retrieval-augmented generation: an AI model answers by first retrieving relevant passages from your own documents, then citing them, rather than guessing from memory.

AI & Machine Learning

Row-level security

Restricting which rows of a table a given user can see, so two people querying the same table get different results based on who they are.

Security & Data Protection

PII

Personally identifiable information: anything that can identify a specific person, the category of data that drives most privacy obligations.

Security & Data Protection

OLTP vs OLAP

OLTP systems handle many small transactions fast (an order being placed); OLAP systems handle large analytical queries over history (a quarterly report). Rarely the same database.

Database Engineering

Still sounds like a foreign language?

Describe the problem in your own words. We will translate it, not the other way around.