Free to read · reviewed by an editor
Practical tutorials, guides and a community for people who work with technology and data.
Trending now
Most readRetrieval-Augmented Generation
How a language model answers from your own documents: split, embed, retrieve, then ask it to answer from what was found. And where this goes wrong.
Data QualityBest practices
Data OrchestrationHow-to
FinOpsBest practices
Learning paths
Follow a subject from the first step to the last. Every lesson says what you will be able to do, and what comes next.
Learning pathIntermediateFrom removing duplicates and window functions to modelling for reporting, cohort analysis, indexes and query plans: the SQL a working analyst or engineer uses every week, each lesson with queries that were run.
Learning pathBeginnerPlan what to measure, tag campaigns consistently, hand data to your tags properly, find out why events go missing, understand attribution and know what privacy law asks of analytics.
Learning pathBeginnerThe everyday safety net of a software project: undo mistakes in Git, get out of the three common Git problems, write tests with pytest and run them on every pull request.
Learning pathBeginnerRead an experiment honestly, judge a classifier fairly and show the result clearly: p-values, an A/B test worked end to end, classification metrics, chart choice and dashboards.
Learning pathIntermediateHow data gets from source to report without surprises: the ETL and ELT choice, a first pipeline, cleaning, quality checks, safe re-runs and scheduling.
Editors' picks
The pieces we would read first.
Write a crontab entry that actually works, give the job a lock so a slow run cannot be joined by the next one, and send its output somewhere you can read.
Grokem Editorial4 min read
CI/CDTutorial· Beginner
One YAML file turns your test suite into a check that runs on every push and pull request, on several Python versions, with dependency caching.
Application SecurityBest practices
Credentials committed to a repository are found fast and abused faster. How to keep them out, find the ones already there, and what to do the moment one leaks.
Measurement PlansArticle
Most tracking projects start by tagging everything. The ones that pay off start with a one-page document that says which decisions the numbers are for.
Topics
Each topic has tutorials, how-to guides and explainers, from the first steps to the details that bite.
Managing, engineering, analysing and presenting data.
Languages, frameworks and the craft of building software.
Query plans, indexes and day-to-day database work.
Cloud platforms and services for data and analytics.
New
DNSExplainerResolvers, root and TLD servers, authoritative servers, caching and TTL, and the record types you will actually edit, explained in the order a lookup happens.
Retrieval-Augmented GenerationExplainerHow a language model answers from your own documents: split, embed, retrieve, then ask it to answer from what was found. And where this goes wrong.
FinOpsBest practicesBudgets with alerts, a tagging scheme, idle resources, sensible storage tiers and a monthly review: the habits that stop a cloud bill from becoming a surprise.
Data OrchestrationHow-toWrite a crontab entry that actually works, give the job a lock so a slow run cannot be joined by the next one, and send its output somewhere you can read.
Topic
Managing, engineering, analysing and presenting data.
Data OrchestrationHow-toWrite a crontab entry that actually works, give the job a lock so a slow run cannot be joined by the next one, and send its output somewhere you can read.
Data QualityBest practicesTen checks that catch most bad data before it reaches a dashboard, with a small runnable example where every check is a query that returns the bad rows.
GDPRExplainerPersonal data, lawful basis, consent for cookies, retention and transfers, explained for people who run analytics. General information, not legal advice.
Measurement PlansArticleMost tracking projects start by tagging everything. The ones that pay off start with a one-page document that says which decisions the numbers are for.
DashboardsBest practicesDashboards fail by showing everything to everyone. Ten checks on purpose, layout, labels and trust that make one people actually use.
Topic
Languages, frameworks and the craft of building software.
Version ControlTroubleshootingThe three messages that stop most people: what 'detached HEAD', 'fetch first' and 'CONFLICT' mean, and the exact commands that get you out of each.
RESTIntegrationCollect every page from an API that sometimes returns 503 or 429: follow the cursor, retry only what is worth retrying, wait with backoff and jitter, and stop when it is hopeless.
RESTBest practicesResources, methods, status codes, errors, pagination, versioning and security: the decisions that make an HTTP API predictable for the people who build on it.
RESTReferenceEvery status code a typical web application or API returns, grouped by what the client should do next, with the mistakes that cause each one.
APIs & IntegrationsComparisonThree ways to expose and call services, compared on data fetching, typing, performance, browser support and tooling, with a short guide to picking one.
Topic
Query plans, indexes and day-to-day database work.
Query OptimizationHow-toAsk the database how it will run a query, read the plan from the inside out, and use it to decide whether an index, a rewrite or nothing at all is the fix.
RelationalExplainerWhat atomicity, consistency, isolation and durability promise, shown with a money transfer that fails half-way and leaves nothing behind.
RelationalTroubleshootingWhy SQLite refuses a second writer, how to reproduce the error in six lines, and the changes (shorter transactions, a timeout, WAL mode) that stop it.
Backup & RecoveryBest practicesA backup you have never restored is a hope. Twelve checks that turn backups into a recovery you can rely on, with the reason behind each.
IndexingExplainerWhat an index is, what it costs, how to read the difference between a scan and a search, and the four reasons a query ignores the index you made.
Community
Questions get answers, the person who asked accepts the one that worked, and new members' first posts are read by a person before they appear.
A script runs every 15 minutes from cron. Last week one run took 25 minutes, the next one started while it was still going, and both wrote to the same file. How do I stop a second…
Every beginner portfolio I see has the same three datasets. What would you put in a portfolio if you were hiring? What makes you stop scrolling, in a good way?
The provider sends a signature header. My framework gives me the parsed JSON by the time my handler runs. If I serialise it again and hash that, the signature never matches. What…
I am loading a few thousand records from a public API in a loop and after a while every request returns 429. I added time.sleep(1) but it still happens. What is the correct way to…
People
The people who answer the most questions.
0 reputation
The people who edit Grokem. We answer when we can and point to the page that says more.
10 replies · 1 question
0 reputation
Data engineer. Opinions about idempotent loads.
8 replies · 0 questions
0 reputation
Python for data, a little DevOps on the side.
4 replies · 2 questions
0 reputation
Marketing analyst, mostly GA4 and spreadsheets.
4 replies · 2 questions
Questions
Short answers to what most people ask on their first visit: reading, accounts, progress, getting help and fixing mistakes.
Anything else? Write to us.
What you can expect
Every piece is approved by an editor before it is published.
Each page shows when it was published, updated and last verified.
Found a mistake or an outdated step? Tell us and we fix it, and say so on the page.