Monitor Your Business Data and Detect Anomalies Using AI

Companies are drenched in business data, and making sense of it all is no simple task these days. Using better tools might help, for example data analytics systems that monitor your data and detect any anomalies automatically. is a company offering this type of tool, and in this blog post, I interview their founder Sean Byrnes.

He has been an entrepreneur for 15 years and is currently the CEO and Founder of Prior to Outlier, Sean was the founder of Flurry, the leader in advertising and analytics services for mobile applications acquired by Yahoo! in 2014.

He is also an advisor, mentor and angel investor in the San Francisco bay area. Sean earned an undergraduate engineering degree (with a focus on robotics) from Dartmouth College and a Master of Engineering in Computer Science (with a focus on artificial intelligence) from Cornell University.

His company provides a very interesting solution to the problem of finding the nail in the haystack.

Here is the interview!

Sean Byrnes, CEO of

What is Your Company’s Background?

While at Flurry, I met with hundreds of companies to discuss how they use data to make decisions. The most common question they would ask me is “What should I look for in all of this data I have?”

I realized that the last ten years of business intelligence innovation had centered on answering questions we already had, while the next ten years would focus on helping us ask better questions.

When I started Outlier in 2015, I didn’t know if it was possible to build a product that would solve this problem. For the first 6 months, my co-founder and I acted as consultants, manually pouring through the data for clients looking for new insights and new questions.

The impact we had on our clients was immense and proved there was value there, but it took us 5 attempts to build a product that was able to truly address the product.

Today, Outlier uses over a dozen forms of statistical machine learning to turn raw business data into automatically generated insights for business users.

As with most AI systems it took us years to refine and tune all of those models to work together to produce a high quality result.

What Problem Does Outlier Solve?

Outlier monitors your business data and notifies you when unexpected changes occur, such as shifts in customer behavior or demographics.

What Does Outlier do, and why is it Important to Your Customers?

What is Outlier?

Outlier is a new way of thinking about business intelligence.

Instead of creating new dashboards or running queries when questions arise, Outlier watches all of your business data for you and lets you know when unexpected changes happen in customer behavior, demographics and conversions.

Why is Outlier necessary?

Automated analysis and detection is critical for modern online businesses who touch millions of consumers every day.

The cost of missing a shift in consumer preference, the effects of competition or problems in your products can be measured in millions of dollars. With so much data being produced from every part of the business, there is no way for humans to look through all of it to stay ahead.

Customers can set Outlier up in a matter of minutes and then start every day with an email waiting in their inbox that has the top 4-5 things going on in their business they need to know.

Outlier Story Example

By tailoring insights to each individual, Outlier is like giving everyone at your company a dedicated business analyst who is constantly looking for hidden problems and opportunities.

What makes Outlier different?

Almost all business intelligence tools available today require companies to consolidate their data into a single data warehouse or data lake and then create complex data models to use to generate charts and dashboards. That process can take months or years to complete, and is expensive to maintain as data grows and changes.

Outlier instead connects to business data wherever it lives, be that a cloud service or a SQL database. It takes only minutes to set up and can find insights across and between many different systems by using statistical machine learning without any input or configuration from the user.

This is how Outlier can fill in the blind spot that most organizations have behind their dashboards, and ask new questions about emerging trends in the business.

How Does Outlier use AI?

Outlier uses over a dozen forms of statistical machine learning in an eight stage processing pipeline that turns raw metrics data into human-readable insights. At a high level, that involves four steps:

  1. Modeling every dimension of all the business data to understand what is normal for every one of millions of customer segments.
  2. Detection of anomalous events that don’t fit the expected behavior of the business.
  3. Clustering of events to find larger, higher level patterns.
  4. Filtering through the clusters of events to select the few that best match the interests of the user.

Most of our customers have millions of dimensions of data, but only the time to absorb 4-5 new insights every day. That means that every step of the Outlier process needs to reduce the universe of possible insights by 10x or more.

Such fidelity requires both multiple learning systems and immense data to train them to recognize the right patterns and filter accordingly.

In Your Mind, What is the Future of AI in Business and Marketing?

The future, which is arriving today, is to automate decision making.

In many organizations, many people’s jobs are to look at inputs (in the form of spreadsheets, charts and reports) and take one of a series of pre-defined actions.

We’ve already seen this with online advertising where humans with spreadsheets have been replaced by DSP which programmatically buy millions of ads every minute.

There are many jobs which today seem to require a human but will be automated in the coming years.

Do You Think Marketers Will be Replaced by AI Robots?

For the most part they have been already. The vast majority of online advertising is purchased programmatically, with AI powered systems buying and selling ad inventory millions of times a minute.

This is a large reason why you see contraction and consolidation in the digital agency market, as humans are becoming less necessary to buy and sell inventory. Buying of non-online advertising (TV, radio, etc) is driven by data and analytics that are using AI to make recommendations.

Today, humans are still at the center of brand marketing and this will likely stay true in markets where creativity is more important than sheer performance. AI is vastly superior to humans in improving performance, but humans still dominate in creativity.

Do You Have Any Other Thoughts on AI in Business and Marketing?

Any task that requires repetitive actions and can be taught to a human in a matter of weeks will be automated using AI systems. This includes everything from accounting to legal research, since AI systems can do them so much more quickly for lower cost.

Areas such as design, strategy and branding will remain the domain of humans until AI systems are build that can be creative and innovative, which will happen eventually.

In this interview, Sean Byrnes gave an interesting view into how monitoring business data can be automated and improved with anomaly detection using artificial intelligence.


I am an author, speaker and consultant in marketing automation and artificial intelligence.

Do you need help with marketing automation or AI-based solutions? Contact me and let’s discuss how I can help you!

Magnus Unemyr

Author, speaker and consultant in the aras of marketing automation, artificial intelligence, and the Internet-Of-Things. Contact me if you need help! Learn more.