The coach who is always on call.
How Learning Tapestry built an AI-powered professional learning coach that gives every educator on-demand access to decades of research-backed teaching strategies.
The Problem
In 2023, ISTE and ASCD completed a historic merger, combining ASCD’s 80 years of professional learning content with ISTE’s mission to accelerate technology innovation in education. The new organization represented the best of both worlds: deep, research-backed teaching knowledge and a forward-looking technology vision.
ASCD’s content library is vast: thousands of publications covering evidence-based teaching strategies, differentiation techniques, assessment practices, and classroom management approaches accumulated over eight decades. The knowledge is there, but getting it to teachers at the moment they needed it was not.
A fourth-grade teacher preparing for tomorrow’s reading lesson at 9 PM on a Sunday couldn’t search through archives to find the right differentiation strategy. A new teacher struggling with classroom management couldn’t quickly surface the relevant chapter from the right ASCD book. The content was locked behind traditional search interfaces that required knowing what to look for before you could find it.
ISTE had already tried to solve this with a previous vendor, but the project was stalled. The clock was ticking. ISTE wanted to announce at ISTELive 23, their flagship conference with 17,500 attendees.
We had decades of research-backed teaching strategies. The challenge was getting the right answer to the right teacher at the right moment.
80 years of content. 170,000 members. No way to search it all at 9 PM on a Sunday.
The Solution
Learning Tapestry was brought in and immediately saw a faster path. While the ed-tech industry was obsessing over custom-trained LLMs and fine-tuning models on proprietary data, they were expensive, slow to build, and unnecessary for this particular use case.
LT proposed something different at a fraction of the cost and time: Retrieval-Augmented Generation (RAG) with vector embeddings over ASCD’s curated content library. Instead of training a custom model, LT suggested that ISTE use an existing frontier LLM and ground its answers in ASCD’s vetted publications.
The previous vendor had been building from the ground up, and LT recognized that the smartest approach was to leverage and stitch together existing tools innovatively that could be demonstrated in weeks, not months.
The core insight was the “walled garden” approach. Unlike general-purpose chatbots trained on the open internet, this AI coach draws exclusively from ASCD’s vetted, research-based publications. Every answer is grounded in content that has been peer-reviewed, edited, and published by one of education’s most respected organizations. This walled garden architecture reflects the emerging regulatory and liability landscape around AI in education, where systems that hallucinate or cite nonexistent sources create real institutional risk for the organizations that deploy them.
And critically, it can cite its sources. When the coach recommends a differentiation strategy, it tells the teacher exactly which ASCD publication the recommendation comes from. This gives teachers a digital trail to verify every recommendation, building trust in the AI by making it transparent.
The Impact
The AI coach was just the beginning. ISTE+ASCD has since become the largest provider of AI professional development for educators, and the approach Learning Tapestry proved, RAG over curated content instead of custom models, has become the industry standard for educational AI applications.
ISTE’s AI initiatives have expanded dramatically. GenerationAI, backed by a $10 million Google.org grant, is building AI literacy programs for students. A partnership with Google aims to train 6 million teachers on using AI effectively in the classroom. The ISTE Seal of Alignment program now evaluates AI products against research-based standards.
The pattern LT established, ground AI in vetted content, cite sources, build trust through transparency, runs through all of it. What started as a professional learning chatbot became a proof of concept for how AI should work in education: not as a replacement for human expertise, but as a way to make decades of accumulated knowledge accessible at the moment of need.
AI will be as transformational to learning as electricity and the internet were.
A teacher at 9 PM on a Sunday can now ask a question and get a research-backed answer in seconds, with a citation to verify it. That’s the difference between having access to a library and having access to a coach who has read the entire library. Learning Tapestry built the coach.
Learning Tapestry delivered a working AI solution in weeks where others had spent months. Their ability to move fast and leverage existing tools was exactly what we needed.
Richard Culatta, CEO, ISTE+ASCD
References
- ISTE, “About ISTE,” iste.org. ISTE and ASCD unified in 2023.
- ISTE, “AI in Education,” iste.org.
- Google.org, “GenerationAI: $10M to build AI literacy,” blog.google.
- ISTE, “AI in Education — Training 6 million teachers,” iste.org.
- Richard Culatta, remarks at ISTELive conference on AI’s transformational potential in education.