Hi! I’m Mats, a Norwegian who speaks seven languages. For the past few years I’ve been working with Google to improve search ads, and built the data insights team from the ground up. A part of my work there was to train both humans and machines to identify key points that could improve ad performance through sophisticated automated pipelines. Then I realized that a lot of companies could benefit from modern data foundations with easy data accessibility, the speed of automated workflows, and smart logic for real-time notifications.
The idea of Secure ML was born. I had the ambition of helping companies step up their data-game and scale with confidence knowing that their data are secured against modern security threats.

Meet the Founder

A portfolio site

Whether it’s to inquire about our solutions, connect, or if you want to discuss a blog post in more detail, we’d love to hear from you! Feel free to reach out on any of our socials, or submit a form to get in touch.

Get in touch!

Secure ML

Data Engineering is the backbone of any data-driven organization. Without a solid data engineering foundation any insights drawn from your data has the underlying uncertainty of whether your data is accurate. Any predictions inferred through machine learning can lead you astray when the underlying data is questionable. We can help you build confidence and trust in your data, and we'll ensure that your data is securely extracted and stored. You should focus on what you do best, let us deal with making your data work for you.

The backbone of any data-driven organization. We can make yours flexible.

Data Engineering

Data Analytics is where actionable insights are created about your product, and customers, and is the domain within data where most organizations will profit the most. By understanding customer behavior you can see what their pain-points are, and take the guesswork out of knowing where your product or service can be improved. We can help build the tools to help you gain a better understanding of your product, and take it to the next level. Analytics is also important for guiding the development of any machine learning services you may want.

Nothing is more dangerous than data that lies. We can make yours honest.

Data Analytics

Machine Learning builds on the data stored through Data Engineering and the metrics and insights derived through Data Analytics to generate predictions on new data points. These predictions open up vast opportunities for automation and efficiency increases. We can help you through the entire data process from engineering to machine learning, and create scalable and secure solutions suitable to your needs.

Automation makes all of our lives a little easier. We can make yours intelligent.

Machine Learning

Cybersecurity is important to any modern organization, and is most effective when woven into the architecture, rather than being added on afterwards. A secure system is one that doesn't rely on users doing the right thing. That being said, it helps a lot, and that's why we have a strong focus on making any security features as frictionless as possible. We can help you to build truly Secure ML.

Cybersecurity systems keep your organization safe. We can make yours invisible.

Cybersecurity


Are you interested in pursuing a career in cybersecurity? With the increasing reliance on technology and the rise of cyber threats, the demand for skilled cybersecurity professionals has never been higher. Here are some steps you can take to get into the field in 2023.

## 1. Educate Yourself

A strong foundation in computer science and information technology is essential for a career in cybersecurity. If you're just starting out, consider earning a degree in a related field, such as computer science, information technology, or cybersecurity. Online courses and certifications can also be helpful in learning the necessary skills and knowledge.

## 2. Gain Experience

In addition to formal education, it's important to gain practical experience in the field. This can be through internships, part-time jobs, or even personal projects. Building a portfolio of projects and demonstrating your skills and knowledge to potential employers will greatly increase your chances of landing a job in cybersecurity.

## 3. Stay Up-to-Date

The field of cybersecurity is constantly evolving, with new technologies and threats emerging all the time. It's important to stay current on the latest developments and best practices in order to be a competitive candidate in the job market. Consider joining professional organizations, attending conferences and workshops, and participating in online communities to stay up-to-date.

## 4. Get Certified

Earning a cybersecurity certification can be a valuable way to demonstrate your expertise and differentiate yourself from other candidates. Popular options include the Certified Information Systems Security Professional (CISSP) and the Certified Ethical Hacker (CEH). Keep in mind that certifications often require a certain level of experience in the field, so it's important to gain some practical experience before attempting to earn a certification.

## 5. Network

Networking is an important part of any job search, and it's especially crucial in the field of cybersecurity. Connecting with other professionals in the field, whether through online communities or in-person events, can help you learn about job openings and make valuable connections that can help you advance your career.

Pursuing a career in cybersecurity can be rewarding and challenging. By educating yourself, gaining experience, staying up-to-date, getting certified, and networking, you'll be well on your way to a successful career in this exciting field.


Breaking into the cybersecurity field can be difficult. Here is some hopefully helpful advice.

How to Get into Cybersecurity in 2023


Data analytics is a rapidly growing field with a wide range of career opportunities. If you're interested in getting into data analytics in 2023, here are some tips to help you get started:

## 1. Learn the basics

To be successful in data analytics, you'll need a strong foundation in math, statistics, and programming. Start by learning basic concepts like probability, statistics, and linear algebra, as well as programming languages like Python and R.

## 2. Build up your portfolio

Having a portfolio of projects and examples of your work is essential for landing a data analytics job. Consider participating in online hackathons or working on open-source projects to gain practical experience and build up your portfolio.

## 3. Get certified

Many companies look for data analysts with formal certifications, such as the Google Analytics Individual Qualification or the Microsoft Certified: Data Analyst Associate. These certifications can help you stand out in the job market and demonstrate your skills and knowledge to potential employers.

## 4. Network with other professionals

Networking is key in any field, and data analytics is no exception. Attend industry events, join online communities or forums, and connect with other professionals to learn about new technologies and trends in the field, as well as potential job opportunities.

## 5. Stay up to date

Data analytics is a rapidly evolving field, and it's important to keep up with the latest technologies and trends. This could involve learning new programming languages or frameworks, or staying up to date with new tools and platforms.

By following these tips, you can position yourself as a competitive candidate in the data analytics job market in 2023. With the right skills, experience, and network, you'll be well on your way to a rewarding career in this exciting field.


Data Analytics is a rapidly growing field. Here are some tips for breaking into your new career.

A Beginner's Guide to Becoming a Data Analyst


Are you interested in a career in data engineering? With the growing demand for skilled professionals in this field, now is a great time to start building your skills and experience. Here are some tips on how to get into data engineering in 2023:

## 1. Learn the basics

Data engineering involves working with large datasets and building systems to process, store, and analyze that data. To be successful in this field, you'll need a strong foundation in programming and data analysis. Start by learning programming languages like Python and Java, as well as SQL and data visualization tools like Tableau.

## 2. Build up your portfolio

Having a portfolio of projects and examples of your work is essential for landing a data engineering job. Consider participating in hackathons or working on open-source projects to gain practical experience and build up your portfolio.

## 3. Get certified

Many companies look for data engineers with formal certifications, such as the Cloudera Certified Data Engineer or the Microsoft Certified: Azure Data Engineer Associate. These certifications can help you stand out in the job market and demonstrate your skills and knowledge to potential employers.

## 4. Network with other professionals

Networking is key in any field, and data engineering is no exception. Attend industry events, join online communities or forums, and connect with other professionals to learn about new technologies and trends in the field, as well as potential job opportunities.

## 5. Stay up to date

Data engineering is a rapidly evolving field, and it's important to keep up with the latest technologies and trends. This could involve learning new programming languages or frameworks, or staying up to date with new tools and platforms.

By following these tips, you can position yourself as a competitive candidate in the data engineering job market in 2023. With the right skills, experience, and network, you'll be well on your way to a rewarding career in this exciting field.


5 Tips for breaking into Data Engineering in 2023

How to Get into Data Engineering in 2023


Are you interested in a career in machine learning? With the growing demand for skilled professionals in this field, now is a great time to start building your skills and experience. Here are some tips on how to get into machine learning in 2023:

## 1. Learn the basics

To get into machine learning, you'll need a strong foundation in math, statistics, and programming. Start by learning basic concepts like probability, statistics, and linear algebra, as well as programming languages like Python and R. Many online resources, such as Coursera and DataCamp, offer courses and tutorials to help you get started.

## 2. Build up your portfolio

Having a portfolio of machine learning projects and examples of your work is essential for landing a job in this field. Consider participating in online hackathons or working on open-source projects to gain practical experience and build up your portfolio. You can also volunteer to work on machine learning projects for non-profit organizations or small businesses in your community.

## 3. Get certified

Many companies look for machine learning professionals with formal certifications, such as the IBM Certified Data Scientist or the Deep Learning Institute Certificate. These certifications can help you stand out in the job market and demonstrate your skills and knowledge to potential employers.

## 4. Network with other professionals

Networking is key in any field, and machine learning is no exception. Attend industry events, join online communities or forums, and connect with other professionals to learn about new technologies and trends in the field, as well as potential job opportunities. You can also reach out to machine learning professionals on LinkedIn to ask for advice and learn about their career path.

## 5. Stay up to date

Machine learning is a rapidly evolving field, and it's important to keep up with the latest technologies and trends. This could involve learning new programming languages or frameworks, or staying up to date with new tools and platforms. Consider subscribing to industry publications or attending conferences and workshops to stay informed about the latest developments in the field.

By following these tips, you can position yourself as a competitive candidate in the machine learning job market in 2023. With the right skills, experience, and network, you'll be well on your way to a rewarding career in this exciting field.


From Stable Diffusion to ChatGPT, machine learning is rapidly growing. Here’s how to get in on the fun!

A Guide for Aspiring Data Scientists in 2023


ChatGPT is a large, state-of-the-art language model that was developed by OpenAI. It is based on the GPT-3 (Generative Pretrained Transformer 3) architecture and is designed to be able to generate human-like text in a variety of different styles and formats.

One of the key features of ChatGPT is its ability to understand and respond to natural language input in a conversational context. This means that it is able to hold a conversation with a human user in much the same way as a person would, making it a valuable tool for applications such as chatbots, customer service agents, and even digital assistants.

The implications of a language model like ChatGPT are significant, particularly in the field of artificial intelligence. As it continues to improve and develop, it has the potential to revolutionize the way we interact with machines, making it easier and more natural for us to communicate with them.

## Potential applications

Here are a few potential applications of ChatGPT:

- **Digital assistants:** Imagine being able to ask a digital assistant for help with a task, and having it not only understand what you want, but also provide you with clear, human-like instructions on how to accomplish it. This could make it easier for people to use technology and take advantage of its many benefits, without needing to be experts in programming or computer science.
- **Customer service:** Another potential application of ChatGPT is in the field of customer service. With its ability to understand and respond to natural language input, it could be used to create chatbots that are able to provide personalized, helpful responses to customer inquiries. This could improve the customer experience and make it easier for businesses to provide high-quality service to their clients.

## Conclusion

Overall, ChatGPT represents a major advance in the field of natural language processing and artificial intelligence. It has the potential to change the way we interact with machines and make it easier for us to use technology in our daily lives. As it continues to develop and improve, it will be interesting to see what other exciting applications and innovations it will enable in the future.


Prompt: Write a blog post about what ChatGPT is and what the implications are for the future.

What is ChatGPT and what are the implications for the future?


I’ve always loved reading books, but sometimes I wished that the book would have a video adaptation and I could just leave it on in the background. I would daydream about having a program where you could pass in a book and it would spit out an animated version of it. Stable Diffusion is a great leap towards that being a reality. Though there is still a long mountain to climb.

In technical terms Stable Diffusion is a latent diffusion model, which converts a text prompt into a latent space representation of the text, and then converts the latent space into pixel space. In more understandable terms, you write a prompt and the model generates an image with a similar latent space as the text. One example is [“a photograph of an astronaut riding a horse”](https://en.wikipedia.org/wiki/Stable_Diffusion#/media/File:A_photograph_of_an_astronaut_riding_a_horse_2022-08-28.png). Another example is the Secure ML logo, which was generated using the prompt “padlock, brain, neurons, brain with padlock, blue, 2d illustration”.

There are many ways you can play around with stable diffusion, and many are trying to learn the art of finetuning their prompts to get the results they want. You can check out [MidJourney](https://midjourney.com/) or [Hugging Face](https://huggingface.co/spaces/stabilityai/stable-diffusion) here to dip your toes into image generation through stable diffusion.


Imagine that you could just tell a computer what you want to see, and the computer creates it. That’s Stable Diffusion.

All you need to know about Stable Diffusion.


Knowing when to use which tool is important, so I’ll start by a quick rundown of each one.

## Tableau

Tableau is a powerful visualization and discovery tool. One aspect of Tableau that sets it apart is its focus on storytelling through a feature it calls Stories. With Stories you can link together dashboards and reports to build a narrative or presentation.

While Tableau has a very point-and-click UI, users are also expected to know SQL which means that this tool falls firmly into the Data field. There is also Tableau Prep which can be used to minimize the need, but as with any of these point-and-click data cleaning solutions it has its limitations.

Tableau has a pretty hefty price tag, and combined with its required skill level, which leads to its user base being mostly composed of companies that already are invested in data. There is a free version, but any visualizations made in it are publicly available.

## Power BI

Power BI is Microsoft’s offering to the data visualization game. Its key selling points are

1. A point-and-click UI which leads to it being easy to use for non-data people. _Though to harness all of Power BI’s power you will need to be able to use DAX_

1. A tight integration to the rest of Microsoft’s productivity suite

If your company already uses Microsoft Office then Power BI will probably be the easiest data viz tool.

Just like Tableau, Power BI also has a pretty hefty price tag, but some of the services may already be included with some Microsoft Office subscriptions.

## Looker Studio

Looker Studio, formerly known as Google Data Studio, is one of Google’s tools for data visualization and dashboard creation. It’s part of the Google Analytics suite, and has integrated connections to Google Ads and other Google Cloud services.

One main selling point for Looker Studio is that it’s free! This makes it an easy choice for budget conscious individuals and companies that don’t necessarily want to publish their visualizations and metrics for everyone to see.

## Conclusion

There is a data visualization tool out there for everyone, perhaps even several. For personal use I use Looker Studio, or the python module [Plotly](https://plotly.com/python/). However, if you’re an organization looking into taking control of your data, I recommend Tableau, even if it can be quite polarizing. It has a steep learning curve, but I feel the storytelling aspect makes your point much clearer. Of course, there are exceptions. One of them being that your organization already is heavily invested in the Microsoft ecosystem. In that case I recommend Power BI. It doesn’t have quite the same polish, but it is a workhorse for business intelligence.


Tableau, Power BI, and Looker Studio are three popular data visualization and business intelligence software. Why and when would you choose one over the other?

Tableau vs Power BI vs Looker Studio


dbt is a very helpful tool that allows some software engineering best practices to be adopted in the data engineering realm. It also helps data teams build reliable and trusted data pipelines. Today I’d like to highlight three aspects where dbt shines.

## Simplifying pipelines

dbt allows analysts and engineers to solely use SQL without needing to branch out to other programming languages. It also allows for smaller, more modular code. Instead of repeating business logic in several places, where inconsistencies might arise, you can write the business logic in one place, and then re-use that component.

## Automated testing

Together with the more modular code dbt also introduces continuous integration into the data pipeline. This means that instead of committing the entire pipeline for every change, components can be individually tested, pushed, and monitored. This allows for a much faster discovery of potential bugs, and fixing said bugs.

## Data alerts

dbt also works as an orchestrator with integrated monitoring. In other words, you can specify schedules for your workflows, and you can set up alerts to notify you about stale data.

## Conclusion

In summary, dbt simplifies data pipelines by introducing software engineering best practices, automated testing, and easy orchestration.


What is dbt (data build tool) and why are all the data engineers talking about it?

What is dbt, and why is it useful?


On September 15th Uber confirmed that it had been infiltrated by a hacker going by the handle Tea Pot, who has ties to the hacker group Lapsus$. How did one individual gain access to many of Uber’s critical systems?

## Social Engineering

As long as passwords are still used, social engineering will be the easiest way for adversaries to gain unauthorized access. More than 90% of data breaches start with a successful phishing campaign. Spear phishing, where the adversary pretends to be a service or person the target trusts, accounts for 60% of these, and in any organization it’s not a matter of if, but when a phishing attack is successful.

After obtaining the login credentials through a successful phishing campaign Tea Pot repeatedly attempted to log in, which repeatedly sent authentication notifications to the compromised employee, and Tea Pot then called this employee posing as an employee working with the authentication service apologizing for the repeated notifications and claiming that the only method they knew of stopping it was to accept the authentication request. And then, they were in.

## Exposed Credentials

While having an intruder sounds bad, it isn’t really that bad as there should be access controls set in place to prevent any, even authorized, individuals from having too much access, and privileged access levels should be tightly monitored. Where this breach took a turn for the worse was that while logged into the corporate intranet through the compromised account, Tea Pot found a script that contained the login credentials for an admin account for Uber’s Privileged Access Management system. From there Tea Pot could take control of many critical services, including the one supposed to detect intrusions.

## Where did Uber go wrong?

It would be naïve to blame the intrusion on the compromised employee. Phishing attacks occur very frequently, and a large organization, such as Uber, should expect accounts to be compromised on a regular basis. The first mistake was to not have alerts for multiple failed logins. As for the hardcoded credentials, people will be lazy and take the path of least resistance. If there is an easy alternative to gaining access to the privileged account, that is what would be used. Secondly, privileged accounts should require more than one person to log in.

You can’t expect people to be perfect, that’s why you design for imperfection.


What went wrong to allow an intruder to infiltrate Uber’s network, and measures could have prevented it?

What can we learn from the Uber breach?


Something has always existed. According to physics, there can never be true physical nothingness—though there can be times when existence resembles nothing, such as a vacuum (the state of minimum possible energy) (Phys.org). Creating a space where there are no quantum fluctuations requires an enormous amount of energy, and there would be a remnant of that energy in that space afterwards if the fluctuations were flushed out, plus an unstable environment (1veritasium). Even on computers, deleted data is not actually tossed away, by rather written over.

# h1 Heading

## h2 Heading

### h3 Heading

#### h4 Heading

##### h5 Heading

###### h6 Heading

## Horizontal Rules

---

## Emphasis

**This is bold text**

**This is bold text**

_This is italic text_

_This is italic text_

~~Strikethrough~~

## Blockquotes

> Blockquotes can also be nested...
>
> > ...by using additional greater-than signs right next to each other...
> >
> > > ...or with spaces between arrows.

## Lists

Unordered

- Create a list by starting a line with `+`, `-`, or `*`
- Sub-lists are made by indenting 2 spaces:
  - Marker character change forces new list start:
    - Ac tristique libero volutpat at
    * Facilisis in pretium nisl aliquet
    - Nulla volutpat aliquam velit
- Very easy!

Ordered

1. Lorem ipsum dolor sit amet
2. Consectetur adipiscing elit
3. Integer molestie lorem at massa

4. You can use sequential numbers...
5. ...or keep all the numbers as `1.`

Start numbering with offset:

57. foo
1. bar

## Code

Inline `code`

Indented code

    // Some comments
    line 1 of code
    line 2 of code
    line 3 of code

Block code "fences"

```
Sample text here...
```

Syntax highlighting

```js
var foo = function (bar) {
  return bar++;
};

console.log(foo(5));
```

## Tables

| Option | Description                                                               |
| ------ | ------------------------------------------------------------------------- |
| data   | path to data files to supply the data that will be passed into templates. |
| engine | engine to be used for processing templates. Handlebars is the default.    |
| ext    | extension to be used for dest files.                                      |

Right aligned columns

| Option |                                                               Description |
| -----: | ------------------------------------------------------------------------: |
|   data | path to data files to supply the data that will be passed into templates. |
| engine |    engine to be used for processing templates. Handlebars is the default. |
|    ext |                                      extension to be used for dest files. |

## Links

[link text](http://dev.nodeca.com)

[link with title](http://nodeca.github.io/pica/demo/ 'title text!')

Autoconverted link https://github.com/nodeca/pica (enable linkify to see)


Something has always existed. According to physics, there can never be true physical nothingness—though there can be times when existence resembles nothing, such as a vacuum (the state of minimum possible energy) (Phys.org). Creating a space where there are no quantum.

Blog