Building a GenAI Roadmap for Educational Content Creation

About NewGlobe

NewGlobe Education is a leader in enhancing global education systems, focusing on inclusive, equitable, and transformative learning. NewGlobe supports national and state governments by creating powerful technology-enabled education systems. An education expert and leader in learning, NewGlobe has unequaled experience in dramatically transforming educational outcomes at speed and at scale. 

NewGlobe’s Challenge

Tim Sullivan, Chief Academic Officer at New Globe, explained that much of the leg work that goes into rolling out an educational system to a new geographical location was centered around content creation, specifically the time and team needed. Additionally, implementing education systems globally required content to be created in many different languages. Sullivan surmised that generative AI might provide NewGlobe with the ability to automate the content creation process and increase the speed of their go-to-market strategy. NewGlobe has engaged Tribe AI to build a proof of concept (POC) aimed at this notion.

“Our value is a function of our scale. We needed a leaner and speedier replication model to increase our impact. Maybe GenAI is the answer?,” contemplated Sullivan.

Why Tribe AI?

NewGlobe knew they needed a partner with extensive experience and understanding of AI, from ideation to experimentation to integration. Through a recommendation from Francisco Partners, a trusted and well-respected advisor, NewGlobe began a POC-centered engagement with Tribe AI. 

Proposed Solution

At the start of the engagement between NewGlobe and Tribe AI, the teams spent much time focused on discovery. Tribe AI’s team knew they needed to fully understand NewGlobe’s business processes in order to recommend where and how GenAI could produce meaningful outcomes. During this strategy phase, it was uncovered that NewGlobe’s team has produced over 100,000 teacher guides. This process is largely manual and a major impediment to the 6-month go-to-market speed that NewGlobe is working to improve. Given the criticality to the business, it was determined that Tribe AI would build a POC aimed at utilizing generative AI to automate the creation of teacher guides. The POC work aligned brilliantly with NewGlobe’s mission to continuously enhance the quality and efficiency of education systems globally. 

Specifically, the collaboration focused on creating a tool that incorporated the following considerations:

  1. Employ an existing AI model to generate text-based teacher guides, ensuring alignment with NewGlobe's educational standards and formats
  1. Aim for compatibility of the AI-generated guides with NewGlobe’s existing content management system, primarily focusing on data export in a compatible format
  1. Conduct a testing phase with NewGlobe to assess the quality and accuracy of the AI-generated content against NewGlobe’s established educational criteria

Technical Details

NewGlobe had previously been using a largely manual process for producing teacher guides, especially those that required bespoke content. The work performed was labor-intensive, at high volume, and within strict timeframes. 

Specifically, the process required an associate to take the content and incorporate it into the teacher guide template. The associate would then add tags for XML publishing and upload the teacher guide to a CRM system that would push the content out to teacher tablets. 

Tribe AI automated much of this content generation process by developing a large language model (LLM) that ingests both the content sheet and the teacher guide template then generates a comprehensive teacher’s guide all at once.

Tech Stack Details

Full-stack cloud-based application working alongside the existing  NewGlobe environment: 

  • Cloud: Amazon Web Services (AWS) Bedrock
  • Large language models: Claude-2
  • Languages used: Javascript React, Python Django
  • Content Management System: Xylene

Developing the Roadmap

During the engagement with Tribe AI and NewGlobe, the outcomes delivered during the POC supported the development of an innovative GenAI roadmap for content creation.

The teacher guide accuracy level remained fairly consistent (rated 7/10 in manual creation & 6/10 for GenAI creation) eliminating some prior concerns surrounding quality. Sullivan expressed satisfaction with the accuracy levels, explaining that likely an additional layer of manual proofreading would be required for GenAI-created teacher guides. He predicted that the savings from using GenAI for content creation would more than offset these additional proofreading expenses.

Although the POC engagement was centered around specific subject content in the English language, its success suggests that the outcomes could be extended into other languages and additional subjects. Sullivan expressed his optimism in how this alone could open the door –more quickly– to new geographical locations for NewGlobe.

Tribe Team Members

Will: Engagement Manager

Ria: Product Manager

Alex: GenAI Lead

Gabriel: Engineer

NewGlobe’s Experience Working with Tribe

Sullivan expressed deep gratitude for the Tribe AI team’s careful and thorough work during the discovery phase.

“Tribe provided a real investment in understanding our business. They served as AI experts evaluating our current processes and challenges. They helped us answer the question ‘where do we think AI could come in and help?’”, said Sullivan.

At the end of the engagement, Sullivan shared confidence in the customized AI roadmap his team developed hand-in-hand with the help of Tribe AI. 

“We now know with confidence that GenAI can further our organization’s impact. And, we’re optimistically looking to uncover other ways we can leverage AI to further innovate the industry,” adds Sullivan.

Impact

The teacher guide creation process had previously required 4 hours of manual labor by an associate. Now, using the LLM developed by Tribe AI, the process would only take 10 minutes. In 2024, alone, NewGlobe has predicted the need to create 30,000 new teacher guides which would require 120,000 hours of manpower. Assuming associates are paid around $7/hour, Tribe AI’s solution would save NewGlobe $835,000, if scaled appropriately. Excluding human review, the total cost to generate a single teacher guide using the LLM would be just $0.15. Using NewGlobe’s previous process, teacher guides were tagged with a cost of $28/each.

“Our organization needed a smart pathway to expansion and that’s just what Tribe AI delivered,” said Sullivan.

The Future

NewGlobe has –since its inception– relied on the intersection of data and technology to improve learning outcomes for more than 2 million kids in 6 countries. But, now they also have a roadmap for utilizing AI to further the organization’s geographical reach. The addition of GenAI also strengthens the organization’s IP value. 

“Using GenAI will open so many doors for us, in regard to optimization and growth. Not to mention, deepen the moat against our competitors," comments Sullivan.

While the content creation process may become more efficient thanks to GenAI, there are some aspects of program expansion, such as teacher training and hiring, that may still require significant time and resources. Even so, Sullivan predicts a significant reduction in the number of overall staff needed for future large-scale expansions, especially in non-English speaking locations. 

Sullivan foresees NewGlobe may also look to GenAI to assist in developing audio coaching for teachers, creating textbooks, building exams, OCR for marking, and more.

Related Case Studies

Case Study

VitalSource Leans on GenAI to Reimagine Content Discoverability for Higher Ed Faculty

Case Study

Togal.ai powers the construction industry into the age of machine learning

Case Study

How Wingspan built a machine learning roadmap with Tribe AI

Case Study

Francisco Partners Accelerates Portfolio AI Efforts with Tribe AI

Case Study

How Nota Built a Roadmap for AI-enabled Journalism with Help from Tribe

Case Study

Building a Proprietary Investment Engine Using Public Data for a Top PE Firm

Case Study

How Tribe AI Built a Model on GCP That Increased Security Questionnaire Auditor Efficiency by 55%

Case Study

Insurance Company Uses ML to Optimize Pricing

Case Study

GenAI Solutions: How Bright Transformed Workforce Training with Tribe AI

Case Study

How Fantasmo is using machine learning to make GPS obsolete

Case Study

Kettle uses machine learning to balance risk in a changing climate

Case Study

How Tribe AI Shaped Truebit’s AI Strategy

Case Study

Taking a Data-Driven Approach to Women's Fertility with Rita

Case Study

How Togal AI Built the World's Fastest Estimation Software on AWS

Case Study

Tribe AI & Venture Labs: Accelerating Startups with Tailored AI Expertise

Case Study

Native Instruments Leverages Amazon Bedrock for Smarter, More Intuitive Search and Discovery for Music Creators

Case Study

Boomi Leverages Amazon Bedrock for Faster Help Desk Responses

Case Study

How Tribe Helped Reservoir Bring Finance Infrastructure to NFT Trading

Tribe helps organizations rapidly deploy AI solutions that have real business impact.

No items found.