Automated Insights Blog

Stay up-to-date with the latest product updates and new advancements from the leader in personalized content.

Read Our New White Paper on Natural Language Generation

Aug 27, 2014


We’ve just published a new white paper, “Big Data Needs Big Insights: The Business Case for Natural Language Generation.”

As businesses spend big on Big Data, there’s a growing need to derive actionable insights.

Our paper explores why Natural Language Generation (NLG) platforms succeed where dashboards and data scientists fail.

"Big Data Needs Big Insights" explains how NLG platforms can deliver real-time, engaging, actionable insights for every level of your business.

Download our white paper for free right now.

Still not convinced you’ve got to read this white paper?

Here’s an excerpt to get you started:

Read More

Does Our Content Sound Robotic? | Ask Ai

Aug 26, 2014

Here’s the latest installment of our “Ask Ai” video series!

In this episode, I test whether our content sounds robotic. Can you tell which piece is automated? Watch and learn!

Got a question? Find and follow us on Twitter @AInsights.

Watch our previous episode to learn more about the “robots” in our office.

If you’ve watched the video and you want to read the content in question, see below.

Read More

50 Data Science Gurus (and 10 Organizations) You Must Follow On Twitter

Aug 18, 2014

Get ready to up your data science game on Twitter.

Our own Ben Bruning (@NCSUalum) recently scoured the Twitterverse for top data science accounts.

Ben started with a search for “data science” and “data scientist,” then chose accounts with a strong focus on data topics.

He also included some popular accounts that occasionally go off-topic but still frequently drop tasty nuggets of data-driven wisdom.

The result: a list of 50 (update: now 51!) excellent data science gurus to follow on Twitter, plus 10 follow-worthy data science organizations.

We’ve named all the accounts at the end of this post. But first, you can read all sixty accounts in one fell swoop with this glorious Twitter list:

Did we leave some amazing data science accounts off this list? Absolutely!

Find and follow us @AInsights to let us know your favorite producers of data science tweets.

UPDATE 8/19/14: We added Kirk Borne to the list after he proved his #DataWrangling bona fides with the words “Lord, I was born a Wranglin’ man.”

And by the way, if you’re looking for Automated Insights employees on Twitter, Ben’s made a list for that, too.

And now, the names of the chosen:

Read More

Vote For Our SXSW Panel!

Aug 14, 2014

Vote to see my session at SXSW 2015!

We proposed an amazing panel for SXSW Interactive 2015. And we need your help to get accepted!

It’s called “When Robots Write The News, What Will Humans Do?” 

Click here to vote for and read all about it!

Here’s a quick preview from the panel description:

In mid-2014, The Associated Press announced it would use technology from Automated Insights to automate corporate earnings stories. The reaction was dramatic, and many wondered: what will happen to the humans?

AP managing editor Lou Ferrara and Automated Insights CEO Robbie Allen are ready to tackle that question.

Lou and Robbie plan to explore:

  • whether automated content will ever express subjective options.
  • how aspiring journalists should prepare for the jobs of the future.
  • how automation will change the way content is consumed.
  • and more!

Thanks for your vote and we’ll see you at SXSW Interactive 2015!

Where Are The Robots? | Ask Ai

Aug 13, 2014

I’m taking your questions in our new video series, Ask Ai!

In this episode, I interview our own Brian Sewell and uncover the truth about the robots that work in our office.

Ask us your questions on Twitter using #askAi, and be sure to follow us @AInsights.

If you’d like to read the articles shown in the video, check them out on Popular Science, Complex, Engadget, and TechCrunch.

I’ll leave you with this picture:


The Life of an Ai Intern

Aug 7, 2014

Each year Automated Insights (Ai) hires exceptional talent for our internship program. This year we have 10 students and recent graduates from local universities and afar. Students from UNC Chapel Hill and Duke are working side by side, surprisingly without conflict. We also have an intern from Washington University and UNC Charlotte. As well as two students from Georgia Tech in Atlanta, and I am a current student at DePaul University in Chicago. So why did we all decide to spend our summers at Automated Insights? I interviewed some of the interns and we all agree that we are learning a lot and are having a great time doing it.


Chris Dieckaus

Duke University Class of 2016

Computer Science Major

Ai Software Development Intern

Q: What have you learned?

A:What have I learned? So much. I’ve learned more this summer than I have in two years of computer science classes. Some new skills include Ruby, Ruby on Rails, Amazon Web Services, JQuery, Angular, HAML, and more.

Q: What has it been like working in a startup atmosphere?

A: I think the startup culture has spoiled me - I can’t imagine working at a non-startup now. Everything is so laid back. There’s a lot of accountability and personal responsibility. You’re expected to get your work done, but you’re not babysat every step of the way.

Garret Dahms

Wingate University Class of 2014

Mathematics Major

Ai Data Analyst Intern

Q: What have you learned?

A: I have learned about managing a schedule, working in teams, upholding responsibilities and being accountable, and in more detail - about how sports statistics (and other statistics) can be automated.

Q: What is your favorite project you have worked on?

A:  This summer I have worked closely with the project management department and I have had chance to work on many interesting proof of products for the sales team. As a college golfer, I was able to help out on some mock-ups for golf related companies and really enjoyed working on something that I can so closely relate too.  

Isaac Presson

UNC-CH May 2014

Business Administration Major

Ai Product Intern

Q: Why did you come to work here?

A: I wanted to work for a startup in the triangle and Ai seemed like the perfect fit. It is a rapidly growing company (as an intern starting in May, I’ve been here longer than three full time employees, how’s that for growth?) I was also interested in the quantitative science behind fantasy football as an avid player for years.

Q: What has it been like working in a startup atmosphere?

A: The culture is awesome. Instead of working in a cubicle, we work in pods, encouraging collaboration and making assistance very accessible. If you need a break from working, we’re encouraged to play a game of foosball, ping pong, or Super Smash Bros. You can even see Robbie, our CEO, partaking in the fun, playing foosball with co-workers. We’re also in the heart of the American Tobacco Campus, allowing us to eat at cool restaurants right next door, and compete against other startups in tournament style games like company dodgeball.


Ai interns, from left Shan, Michael, Jeff, Isaac and Brooks after winning the American Tobacco Campus Dodgeball tournament. Team photographer, Garret. 

Jeff Chesnutt

Georgia Tech, December 2014

Industrial Engineering Major

Ai Development Intern

Q: Why did you come to work here?

A: I was very interested in getting into the Software development side of engineering and felt that Automated Insights would be a good fit. The atmosphere has been great and the work has been very challenging and thought provoking.

Q: What is your favorite project you have worked on?

A: I really enjoyed working on the first project I was given here. I was told to create a script that would generate a report on memory usage within S3 storage that would ultimately show where Automated Insights was holding old or no longer relevant files.


Interns Shan, Garret, Jeff, and Chris working in their pods

Shan Sullivan

Georgia Tech Class of December 2014

Business Administration Major

Data Analyst Intern

Q: What has it been like working in a startup atmosphere?

A: Not many offices have ping pong, foosball, free food, and a great view of the Durham Bulls baseball field. But aside from the materialistic benefits, the company is very flexible when pertaining to employee preferences and schedules. Because people here are always working hard and everyone knows what their job is, you get the ability to take vacation when you want it, work from home if you aren’t feeling your best, and choose to stay late or get to the office early to do what you have to get done. Since it is such a collaborative environment, everyone knows each others’ responsibilities and are accountable for their own job, which makes the projects very efficient and successful.

Q: What is your favorite project you have worked on?

A: Last summer, I was interning for the first Yahoo! Fantasy Football project, where we created over 300 million personalized recaps for every Yahoo! Fantasy Football matchup and draft. It was extremely rewarding to see all the hard work we put in get such positive reviews, and my friends in my fantasy football league couldn’t believe I actually got to work on something that cool that we all used.

Brooks Stoioff

UNC-CH, Dec 2014

History Major

Prod Team Intern/Quality Assurance

Q:Would you recommend the experience?

A:Without a doubt.  I really cannot put into words what this experience has meant to me.  Education is very important, but I would recommend for any young person to find an internship and get exposure in the workforce.  A classroom can only teach you so much about reality and getting hands-on experience is strongly recommended.  I’ve gotten to watch this company grow exponentially over the past few years and it has been remarkable.  

Many of our interns have worked at Ai for consecutive summers and Shan has brought friends from Georgia Tech to the internship the past two summers. So yes we would recommend this to a friend and we already have.

So what are you waiting for, shoot us an email at we are hiring data interns for the fall.

“A Mind-Boggling Pace”: Futurist Gerd Leonhard’s 2020 Vision

Aug 4, 2014


Gerd Leonhard is an accidental futurist.

After writing a book called “The Future of Music,” people started seeking his views on the future of media, marketing, telecommunications, technology, and more.

Gerd offers foresights on the next five years.

“I call this the immediate future because what we see in 2020 is already here today,” he says.

“All we need to do is imagine how it would embody itself in the future.”

In our interview, Gerd discusses:

  • Why science fiction movies like Her, Oblivion, and Transcendence illustrate near-term realities.
  • Why machine efficiency means a human return to the right brain.
  • Why it’s important to be a “nowist” rather than a futurist.

Listen to the interview above and read some lightly-edited excerpts of Gerd’s thoughts below.

On artificial intelligence, smart machines, and machine learning.

We have this technology already here today using Google Now or simple recommendation engines or what have you. 

So A.I. in a simple way is already there and machines are becoming smarter and there’s been the job report about automation.

People aren’t realizing that this technology is developing at a mind-boggling pace.

So in five years we’re going to face very existential issues about which jobs are taken by machines for what reasons, what we end up doing, whether a machine should be able to independently learn and become self learning, whether language is going to be replaced by automated translation.

All these issues are going to be magnified in a very large way.

Looking at the movies like Her or like Oblivion or of course Transcendence, we are not realizing – we’re talking about our immediate future here. So this is a major topic for the next five years and of course beyond that.

On whether smart machines mean sentient machines.

I don’t mean conscious machines like Singularity point machines.

I’m talking about essentially machines that could emulate human activities to such a degree that we would consider them extremely smart and that they can predict things based on data but not of course be self-aware in the sense of making decisions or changing rules that they were programmed to do.

But essentially machines that will take over a lot of jobs that we used to do that were sort of repetitive like bookkeeping, checkout clerks, navigation, translation. 

That’s all going to have great impact on business models, it’s going to change the manufacturing sector, it’s going to change publishing and media. 

We call this sort of basic AI and not really strong AI in a sense of being aware or being smarter than humans.

On how efficient machines will change human society and activities.

Digital hyper-efficiency is becoming a major economic force.

Reducing the value of transactions, for example, making it easier for us to buy things, but also creating more pain for the provider. So now it’s become sort of seller beware not buyer beware. 

Everything becomes more efficient – Netflix costs $8 a month while a DVD used to cost $25.  This is the hyper-efficiency. This kind of Digital Darwinism in a way, you could say.

What that means for us as people - we’re going to have to figure out where we create the extra human value. Of course we know what this, but we’re going to have to focus on this in a much larger way.

I call this the return to the right brain – creativity, imagination. Basically unorthodoxy. Questioning things, negotiating, recognizing patterns, reading between the lines.

All of which were skills that we were asked to not have in the previous iteration of business because we were supposed to be like a good robot basically and just do our job.

And now we’re going back to saying, “the job can be done by somebody more efficient.”

Humans aren’t efficient but they are inventive. They further things in a different way than an algorithm. I call this the “humarithm,” which is the opposite of the algorithm.

It’s basically an interdependent relationship between the two, so the more algorithms we have the more humarithms is what we need

And this will be our future work. We’ll focus on these kind of things that are “humarithmic,” you could say, based on human qualities.

On a move toward “sustainable capitalism.”

We’re going to an economic system that I would call, others have called, sustainable capitalism or natural capitalism, which will completely rethink the stock market, for example. And that’s a five- to ten-year horizon.

And companies like Unilever and others are already pushing this agenda. And for them sustainable is the new profitable, that’s kind of the headline of that.

On being a “nowist” rather than a futurist.

It’s kind of important now to become what I call more of a “nowist” in a way than a futurist. Because a lot of the things that we’re looking at, they sound like science fiction to some people.

But if you’re looking at what is actually happening, pretty much across all those sectors, we’re at the pivot point of the takeoff of a lot of these ideas right now.

* * *

Follow @AInsights for updates on our company, our technology, and other cool stuff from the future.

Listen to our previous audio interviews:

Adapt to the Future: Pew’s Lee Rainie on Brains, Automation, and the Internet

Futurist Marcel Bullinga Predicts “Waste Education”

The NFL Provides a Quantum Leap for Automated Journalism

Aug 1, 2014


Yet another huge advancement for automated journalism was announced yesterday. Not surprisingly, it came from a vertical that’s been at the forefront of automation technology for years: Professional sports, or in this case specifically, the NFL. 

The NFL announced that it will expand an existing stats-tracking test program and, for the 2014/15 season, will be equipping every player with a sensor under each shoulder pad. The sensors will provide near-real-time information on each player’s location and speed.

Back in May, I gave a talk on the future of automated journalism at the Tow Center for Digital Journalism at Columbia University. During that talk, I devoted some time to discussing the Robot Reporter, a growing network of chips and sensors that collect and deliver data to automated content platforms like Automated Insights’ Wordsmith, which then instantly creates news articles from that data.

One example I gave was Quakebot, the LA Times template-driven software that broke its first widely-recognized earthquake news back in March. The second example was about sports, and all of the sensors currently being used to track events like balls and strikes, measurements like first downs, and the NFL’s existing trial with player sensors.

With those sensors in place, I said, it’s easy for us to make the jump to more qualitative analysis of a game, not just a statistical overview.

That part was met with a lot of excitement, and I spent most of the Q&A talking about whether or not it was true and how big an advancement those sensors were.

In terms of sports reporting, it’s the biggest step forward in automated journalism since play-by-play data.

Traditional box score data is usually limited to the overall game stats for each player and some basic team stats. This handcuffs us in a number of ways. The first and most obvious is that data is usually only available after the conclusion of the game, and even then the delivery is delayed from a few minutes up to a few hours. 

The more damning issue, however, is that it limits us to discussing only broad components of the game. For example, we can sometimes tell if the game was a fourth quarter comeback or if a player contributed greatly to the win. But notice I didn’t say we could call out a player as the reason for the win.

Even if it’s true 99% of the time that a quarterback with 400 yards and three touchdowns was the hero, that 1% of the time when it isn’t requires a lot of hedging of bets, so to speak.

In Automated Insights’ existing sports reporting, we’re already using more advanced play-by-play data, which is compiled by humans using a templated “drop-down” approach and delivered within a few seconds of each play. 

Example: PLAYER A did ACTION B which produced RESULT C. 

Or: Adrian Peterson rushed down the middle for a gain for 4 yards.

That was evolutionary. 

We generate recaps for fantasy football (leagues are forming now!), so we’re die-hard fantasy players, and most of us have watched at least one NFL game by following this play-by-play data in not-so-near-real-time.

Play-by-play might seem like a small addition at first blush, but it gives our content an exponential increase in granularity by taking a lot of the aforementioned guesswork out of our analysis.

Now we can report, with 100% accuracy, that “Adrian Peterson was the hero for the Vikings, with 138 yards rushing on 21 carries, the most important of which was a 23-yard touchdown run up the middle with 0:08 remaining in the game.”

Furthermore, we can summarize and aggregate play-by-play data ourselves, at any time, to create a perpetually updated live box score. And when we access historical data and league data, we’re able to create products like Real Time Insights for Major League Baseball, which we debuted at the OnDeck Sports Technology conference in 2013. 


This new player sensor data, when it becomes available, will provide an extraordinary new dimension to our capabilities, allowing us, for the first time, to actually be able to call a game live.

"On 3rd and 6, the Bucs line up in a pass-protect defense. Adrian Peterson runs left from the 23 yard line. He breaks a tackle at the 15 and finds another gear, sprinting to the left corner of the end zone. Touchdown Vikings. Minnesota goes up 27-24 with the extra point to come and 0:08 seconds left in the game. That was Peterson’s longest touchdown run since Week 2 and, if the score holds, the eighth game-winning touchdown of his career. It was also Darrelle Revis’s 5th missed tackle of the game, a season high.”

We can produce that type of narrative today as quickly as the data is produced. 

And it isn’t too hard to imagine those sensors on: 

Police cars: “A high-speed chase is underway on Interstate 40 just east of Raleigh.”

Commercial airliners: “United Flight 100 bound for Dallas made an emergency landing in Atlanta at 10:20 a.m.”

This new form of journalism can be a big plus for human journalists, whether it’s a heads-up on breaking news like Quakebot or a compilation of vast amounts of data into usable information, much like what we do when we automate earnings reports for the Associated Press.

This is only the beginning for the Robot Reporter, but the next evolution isn’t too far down the road.

(top image via Zebra Technologies website)

Joe Procopio is VP: Product at Automated Insights. He is a serial entrepreneur and a writer. Follow him @jproco and read him at

Adapt to the Future: Pew’s Lee Rainie on Brains, Automation, and the Internet

Jul 25, 2014


Lee Rainie directs the Pew Research Center’s Internet & American Life Project. Its most recent report, Net Threats, came out earlier this month.

Since 2004, Lee has been canvassing leading experts that think about the future; he says he wakes up every day “excited about who I’m going to engage next.”

Below are three fascinating audio segments from my chat with Lee.

Lightly-edited transcripts are below each segment.

On the future of the human brain

There are two times in these surveys that we’ve asked questions about where is the human brain going to be and where are human capacities going to lie – sort of in the artificial intelligence realm.

And what you get in answers like that is, we’re going to keep adapting.

The great story about all species, including the human species, is: we adapt to the environment.

And so all of a sudden when you don’t necessarily have to cram lots of facts into your head to navigate the world because you can easily access them with a swipe of your finger or even now an utterance of your voice, that’ll change our processing capacity in our brains about what we have as memories, what we do with critical thinking and stuff like that.

So many of the evolutionary advantages in the future will be to people who don’t necessarily have lots of facts in their head but have a lot of capacity to do critical thinking and are discerning searchers. They can figure out relatively quickly and well the difference between highly credible information and highly suspect information.

People who do pattern recognition – in this world of hyper-abundant informant, if you can sort through the messiness of it and see things going on relatively quickly, you’ve got an advantage over the slower folks who get lost along the way.

So there will be lots of ways that our thinking capacity, our memory mechanisms, and lots of other ways about thinking and consciousness that will adapt to an environment where we’ve all got an extra lobe of our brain now in the cloud. In an environment where we’ve got data swimming around us all the time that we can access instantaneously.

On labor-saving automation

That is exactly the grand thought that Clay Shirky talks about with cognitive surplus. This stuff is freeing up our culture and our species in so many ways.

Some people will get lost by the wayside. Some people will amuse themselves to death in the Neil Postman sense of things.

But other people are going to have all of this new capacity and free time, in some respects, to be better than they might otherwise have been.

I used to be a reporter. Some of the most unhappy days of my life as a reporter was when I was covering the White House for a couple of months subbing in for another colleague of mine. I hated it.

I worked at the New York Daily News, so I wasn’t quite the A-list publications like The New York Times or The Washington Post. I wasn’t given leaks all the time. And so I became sort of like a stenographer, like dozens of other incredibly thoughtful and incredibly well paid people from other news organizations.

I couldn’t have been happier when I returned to my beat in Congress because then I could roam where I wanted to roam and ask the questions that I wanted to ask of the people I wanted to ask them of.

People were a lot freer with information in this place where 435 little nodes had their information to disseminate rather than one central node.

It’s a little, small example from my life of this larger reality – when stuff that can be automated or stuff that can be relatively easily understood and regurgitated in digital form gets automated, we are freed up to do bigger, better, smarter things.

On the increasing prevalence of the Internet

There’s a long-standing trope that technology becomes its most important when its at its most invisible.

The Internet obviously will become  embedded in so many things that people won’t be aware of it in the days to come. Kind of the way that people aren’t aware of electricity now; the only time they think about it is when it’s denied to them.

* * *

Follow @AInsights for updates on our company, our technology, and other cool stuff from the future.

You may enjoy these other posts:

Big Data Has a Long Way to Go

[AUDIO] Futurist Marcel Bullinga Predicts “Waste Education”

Learn how Wordsmith can help your company

- -

About Automated Insights

Automated Insights is the market leader in producing and publishing engaging personalized content at scale. Learn More