How Albert Scales Educational Content Creation with Automation

Albert is a practice and assessment consolidation platform for 5th through 12th grades that helps teachers personalize their students' learning and improve exam performance. Originally focused on AP test preparation, Albert has expanded to cover SAT, ACT, state exams and other curricula across all subjects. The platform includes sophisticated data tools that help teachers identify where students excel and where they need additional support, creating a virtuous cycle of assessment, instruction, and data.

Albert saves teachers countless hours by providing ready-made homework assignments, quizzes, and assessments, instead of requiring educators to create materials from scratch. Their massive database of questions is designed, vetted, and published by an 11-person content team, who also work with hundreds of freelance subject matter experts. In a very real way, Albert's ability to provide value to their customers relies on having top-notch content: their challenge is to figure out how to scale content creation, all while maintaining quality and accuracy. 

Why Albert Chose Gumloop

As Albert's VP of Academics, David Phelps was searching for a way to scale AI capabilities across his teams in a way that felt accessible to non-technical people.

"Everyone that I manage are former classroom teachers, none of whom have degrees in engineering or any kind of major technical skills," Phelps says. So when he saw a Twitter post showing how simple it was to build automated workflows in Gumloop, he was intrigued: he'd found a tool he could use that wouldn't require technical knowledge — and that, in his words, would "feel like Legos" to use.

He started experimenting, and immediately started having ideas of all the ways in which automation could help Albert. "It scaled quite rapidly once I had a little time to play with Gumloop and get the vision of all of the things we could do with it," Phelps says. "I wanted to find something that was easy to train on, easy to use, could scale with us quickly, had great customer support, and had reasonable pricing. The ROI was super clear to me."

How Albert Uses Gumloop

So far, Albert has implemented two major Gumloop initiatives that have transformed their content operations:

Large-Scale Quality Assurance

For their first major use case, Albert was looking to tackle a critical challenge: maintaining a consistent standard of premium quality across hundreds of thousands of practice questions, accumulated over ten years of operation. Questions might have typos or misspellings; they might be too wordy or too concise. Furthermore, their internal review processes and standards have evolved over the last decade, meaning that what might have been acceptable in the past needs to be updated today.

"We are ultimately in the business of trust," emphasizes Phelps. "When schools and districts give us their money from increasingly limited budgets, every dollar they spend with us is a reflection of trust in the quality and reliability of our content. Even one tiny typo is too many." But how to identify which of their hundreds of thousands of questions might potentially need updates? Checking questions manually at such a scale is functionally impossible — or would be, at the very least, prohibitively expensive and time-consuming.

Albert’s solution was to use Gumloop to build sophisticated error detection workflows that identify questions with issues. They use a Google Sheets integration combined with custom nodes and SQL processing that routes flagged content to human experts for final review and correction.

"With Gumloop, we could finally, at scale, detect small things we wanted to improve across a huge base of questions," Phelps says.

Adaptive Content Generation

After the success of using Gumloop for error detection, the Albert team looked toward an even bigger project: building out their new Albert Adaptive learning suite for math and grammar. This free product offering automatically adjusts question difficulty in real time and assesses student proficiency across thousands of skills, meeting post-COVID demand for more automated educational technology. Additionally, Albert Adaptive helps school districts save money and reduce overhead by consolidating multiple single-subject tools onto one school-wide solution: the long-time dream of Albert's Founder and CEO, Luke Liu.

Of course, building out an entirely new learning suite means creating new questions — and lots of them.

Albert developed a Gumloop workflow that leverages their academic expertise and standards to generate questions for each skill level. Before Gumloop, it would have taken their entire 11-person content team, working full-time with freelancers, more than two years to write, edit, and publish 77,000 new questions.

With Gumloop? They only needed two content team members. And it took those two content team members just four months.

Gumloop’s Impact, by the Numbers

• 180,000: questions quality confirmed by the Albert team with Gumloop
• 77,000: new questions created by the Albert team with Gumloop
• 6x: increase in efficiency per content team member

“I've been nothing but impressed with Gumloop's product and how fast it improves. I've worked with a lot of vendors over my years at Albert and I've never had that kind of responsiveness or relationship with the product.” — David Phelps, VP of Academics, Albert

Albert 🤝 Gumloop

For educational technology companies like Albert, maintaining quality while scaling content creation presents unique challenges. Traditional manual processes simply cannot keep pace with the demand for personalized, adaptive learning experiences that today's educators expect.

Through Gumloop's AI automation platform, Albert has solved what seemed like an impossible equation: maintaining their renowned quality standards while dramatically expanding their content library and capabilities. The result is a more competitive platform that continues to reduce district costs and save teachers time while providing students with better learning experiences.

‍