Domains and owners first. The rest is hypotheses.
Get started with data governance: start with one domain, one owner and one or two critical data products.
We share professional opinions about data on our own data blog
Thoughts on data. Written by people at Glitni.
Glitni is a DN Gazelle 2026
Glitni is a DN Gazelle 2026. On our goal from 2021, our growth and what we still need to improve.
How agentic AI changes data roles
Which roles grow when AI agents write the code, and who in the data team needs upskilling or reskilling?
The data team's harness and context matter more than the model
The harness and context around an AI agent for data decide its answers. How to own and maintain the context.
Apache Ossie shows where your metrics are defined
We tested Apache Ossie, the standard for portable metrics: good for mapping, but immature.
Metadata and ontology have become the foundation AI reads from
Measurements from 2026 show that descriptions, business rules and owners decide whether AI agents answer correctly on your data. How to build the layer.
Data quality: how good does the data need to be, and who decides?
How good the data needs to be depends on what it will be used for. How to get started, from master data to ownership.
How to create value from data: four phases, four pitfalls
Data creates value when the organisation matures, not when a platform is bought. Four phases, four pitfalls.
Glitni | Summary of the First Half of 2026
Glitni's honest summary of the first half of 2026: fully booked, new clients, 16 employees and the framework we gave away.
Data Governance | Data ownership: who owns your data, and what does it cost not to know?
Data ownership is the first step to good data governance. Learn what a data owner does, which mistakes are most common, and how to get started in six concrete steps.
Data Team: You Don't Know What You're Building
Speed doesn't kill data quality. Silent assumptions about what was built and what it's supposed to be used for do. We explore the System 1 and System 2 tension inside data teams.
AI makes us faster. That's not the problem.
How Glitni uses AI in data platform and data engineering work — with clear security boundaries, concrete examples and an assessment of effects and practices.
Product forum, 100 days and Learn more
Product forum for data products: decisions, backlog and change practice. 100-day plan for data product owners. Reading list on data mesh and data contracts.
Change, versioning and deprecation practice
Breaking changes in data products (semantics, keys, time logic), SemVer as a shared language, change policy with notification and transition, and tidy deprecation practice.
Data products as the foundation for AI
Data products are the foundation for AI agents, not just for human sharing. The difference between analytics-ready and AI-ready, nine maturity dimensions, and what must be coded in when the consumer is no longer an analyst.
Quality and reliability
Data quality in data products: exploration/standard/business-critical, SLOs for freshness/availability/quality, code quality as the third dimension, and response patterns for deviations.
In practice: product page, catalogue and components
How data products should look in practice: product page, point of contact, changelog, runbook, catalogue and minimum practice for components.
Interfaces first: sharing and data contracts
How to share data products within a domain, across domains and externally. Data contracts when risk and dependency warrant them.
Value: why data products are worth the effort
Why data products deliver value: less fragmentation, more reuse and measurable benefit. Examples of value hypotheses, metrics and common traps to avoid.
Business canvas and MVDP
Business canvas and Minimum Viable Data Product (MVDP): how to define customers, value and product surface - with Definition of Done and what you can safely postpone.
Rulebook and portfolio
Ten rules that separate data product from component, plus lifecycle, portfolio management and data ownership. Rulebook for data teams that want to keep the term useful.
What is a data product?
What is a data product - and what separates it from a table? Definition, four characteristics and examples of deliverables that deserve product status.
Five Architecture Choices to Make Before Your Data and Analytics Renewal Begins
Five architecture choices that should be made before building starts -- to reduce risk, avoid technical debt and increase delivery speed in data and analytics programs.
Glitni | Summary of the Second Half of 2025
Read our honest summary of the second half of 2025 at Glitni. We look back at new clients - and a new office.
Data Mesh | A Complete Guide to the Four Principles
Data mesh is not technology — it is a framework for data organization based on domain ownership, data as a product, federated governance and self-serve. Learn what it requires in practice.
How a community of practice can regain control of data engineering standards
How do you go from the wild west of data pipelines to reusable factory patterns — without undermining decentralised product teams? In this article, we look at how a small community of practice can coordinate shared learning and take responsibility for standards, patterns and tools for data engineering.
Integration platform and data platform: what is the difference?
The integration platform makes the organisation react, the data platform makes it learn. How to tell them apart.
Declarative data platforms: describe what the data should be, not how
Declarative data platforms let you describe what should be true about the data. How it works in dbt, Spark and Snowflake, and why it suits AI agents.
Glitni | Summary of the First Half of 2025
Read our honest summary of the first half of 2025 at Glitni. We look back on new clients and more new hires.
Guide to Data Contracts
What a data contract is, what it contains, and how to introduce it with ODCS, dbt and one team at a time.
Microsoft Fabric | Glitni Has Microsoft Expertise
Microsoft Fabric | Glitni has a range of consultants with experience in Microsoft Fabric.
Microsoft Fabric | How Glitni Can Help You
Microsoft Fabric | Glitni has a range of consultants with experience in Microsoft Fabric.
Are you struggling with slow or expensive API loading jobs in Azure Data Factory?
We propose a solution to reduce runtime and cost by replacing Copy activities with Azure Functions in Azure Data Factory, which is particularly relevant when dealing with a large number of simple activities.
Glitni | Summary of the Second Half of 2024
Read our honest summary of the second half of 2024 at Glitni. We look back on new clients and more new hires.
How to use unit tests in dbt for macros
Glitni shares a proposed testing framework that enables unit testing of macros in dbt.
Data modelling in the age of AI
AI can build the data model and use it, but it cannot decide what the model should mean. How AI agents are changing data modelling, and what they need from the model.
Generative AI - from talk to action
Discover how to get started quickly with generative AI based on advice from experienced data scientists and data engineers.
Glitni | Summary of the First Half of 2024
Read our honest summary of the first half of 2024 at Glitni. We look back on a stable project portfolio, recruitment and increased visibility.
5 Steps to Get Started with AI Products
Understand how AI can solve cognitive tasks. We guide you through five steps from strategy to implementation, with a focus on Norwegian examples and practical tips.
Data Platform | The Components of a Data Platform
Learn about the components of a modern data platform, from data sources to consumption. Explore how they work together to collect, store, process, and present data.
The Data Warehouse Monolith
Learn more about the transition from data warehouse monoliths to domain-divided data platforms, and the benefits this provides.
Data terms and concepts
Explore the evolution of data-related terms from Business Intelligence to Data Mesh and Microsoft Fabric in this article. Perfect for those who are new to the world of data and want to learn more.
Guest Post: The Value of Metadata
Many data teams take a reactive approach to projects and are often disconnected from the rest of the company. By actively leveraging metadata, they can work much more proactively and focus more on work that delivers maximum value to the company.
Glitni launches the podcast Datautforskerne
New podcast in Norwegian! "Datautforskerne" explores trends, technology and strategies in data. Subscribe to learn about data-driven organisations and get professional tips.
What Does the Tech Stack at Glitni Look Like?
Discover Glitni's tech stack! From Google Cloud to GitHub, see which tools we use for infrastructure, sales, marketing, admin, HR, and development.
Trends | The Landscape for Machine Learning, AI and Data (MAD) - 2024
Explore the trends in machine learning, AI and data architecture for 2024. Key players, technology development and predictions from FirstMark Capital.
Glitni | Summary of the Second Half of 2023
Glitni has had an eventful second half of 2023, with highs and lows along the way. We run another retro for the past half-year.
10 consultant mistakes that deserve a wedgie
Experienced consultants in data and analytics share ten common mistakes. Read to learn, whether you are a consultant or you hire consultants.
Glitni | Summary of the First Half of 2023
The consulting company Glitni is growing, with some friction, but mostly proud moments. We run a retro for the past half-year.
Data Platform | Can it be built in 6 weeks?
Establishing a data platform does not have to take months and years. We outline some simple steps to deliver value quickly -- but remember that development must not stop there.
Tips | 15 books to check out on data and analytics
Learn more about the books you should read on data and analytics
A Complete Guide to DataOps
DataOps makes data delivery fast and reliable: automation, CI/CD, testing and observability. A guide with 10 steps and how it differs from DevOps.
Azure Data Factory's orchestration problem
We have written an article about a key limitation in Azure Data Factory that is not yet resolved by Microsoft Fabric
Synapse Analytics | Glitni Has Azure Expertise
Azure | Glitni has a number of consultants with certifications and experience with Microsoft Azure.
Synapse Analytics | How Glitni Can Help You
Azure Synapse Analytics | Glitni has a number of consultants with certifications and experience with Azure Analytics.
Snowflake | Glitni is a Snowflake Partner
Snowflake | Glitni has experienced consultants who help you with architecture and implementation of Snowflake.
Google BigQuery | Glitni is a Google Partner
Glitni has experienced consultants who can help you with architecture and implementation on Google Cloud.
Databricks | Glitni is a Databricks Partner
Databricks | Glitni has experienced consultants who help you with architecture and implementation of Databricks.
Snowflake | How Glitni Can Help You
Snowflake | Glitni has experienced consultants who help you with architecture and implementation of Snowflake
Google BigQuery | How Glitni Can Help You
Glitni has experienced consultants who can help you with architecture and implementation of Google BigQuery
Databricks | How Glitni Can Help You
Databricks | Glitni has experienced consultants who help you with architecture and implementation of Databricks
GUI vs code: what should I choose for data transformation?
Graphical ETL tools where the logic only exists as clicks are a dead end when AI agents write the code.
Trends | The Landscape for Machine Learning, AI and Data (MAD) - 2023
Glitni summarises the MAD Landscape 2023, which maps the technology development in machine learning, AI and data.
Trends | 7 predictions for data and analytics in 2023
Do you have a full overview of everything needed to become more data-driven? Here are 7 trends you should know about.
Glitni | Summary of the Second Half of 2022 for Glitni
OK. It took some effort to secure projects for everyone. Decision-making processes take time, after all. But it ended well.
Data platform: 5 things to consider before the data platform is ready
Five things the organisation needs in place before a new data platform is ready: strategy, ways of working, change, skills and communication.
Glitni | We enter partnership agreements
Glitni has now entered partnerships with several of the biggest names in modern data platforms: dbt Labs, Snowflake, Google Cloud and Databricks.
Data strategy: what it is, and what it must answer for AI
What a data strategy is, the problem it solves, what it contains, and what AI needs from the layers beneath.
DataOps | Making the Case for a Central Data Platform Team
Cloud-based architecture changes how we work and organize. Here is why you need a centralized platform team.
dbt | Summary of Coalesce 2022
We take a quick look at the launches from dbt Labs during Coalesce 2022, which included a new semantic layer and support for Python-developed models.
Data Platform | Historical Development
We started with reporting on top of databases, moved on to 30 years of data warehousing, before data lake and now data lakehouse have taken over.
Data Lakehouse | What is a Data Lakehouse?
The data lakehouse is the lovechild of the data lake and the data warehouse, suited for storing and processing all kinds of data for reporting and analytics.
Data Warehouse | What is a Data Warehouse?
A data warehouse brings structured data from many sources together in one modelled form with history. How it works in the cloud, and why AI agents need it.
Data Lake | What is a Data Lake?
A data lake is suited for storing all forms of data for analytical user stories, including data we are not entirely sure we will use.
Database | What is a Database?
A database is designed to store data and is a common component in a great many IT solutions, because how else would we be able to keep track of everything that happens?
Glitni | Summary of the First Half of 2022 for Glitni
It has been fun running a consulting company this half-year! Because it is quite fun when everyone is running in the same direction and rolling up their sleeves when it counts.
Data Platform | Get Control of Cloud Costs!
Data processing costs can easily spiral out of control in the cloud. We have some simple tips - you wouldn't want to give away money when it's easy to do something about it, would you?
Analysing JSON in Synapse, BigQuery, Databricks and Snowflake
We have published a Medium article comparing schema-on-read support for JSON across a selection of popular data processing engines
Glitni | Here is What We Learned About Starting a Consulting Company
We have just been through some exercises that we believe most people thinking about starting a consulting company will need to go through, so we are sharing!
Trends | 7 predictions for data and analytics in 2022
Do you have a full overview of everything needed to become data-driven? Here are 7 trends, including data lakehouse, dataops, ELT and Pub/Sub, business analytics, data mesh.
dbt | 10 tips for getting started
Get tips on how to start a dbt project (data build tool) from a data platform engineer. dbt should be considered a central part of any data stack in 2022.
Five places data creates value, and what AI has changed
Five places data creates value in an organisation, two tracks for capturing it, and what generative AI has changed since 2022. With an e-commerce case.
A Complete Guide to DataOps
DataOps makes data delivery fast and reliable: automation, CI/CD, testing and observability. A guide with 10 steps and how it differs from DevOps.