Context
EdDen was an AI lesson-planning tool for K-12 teachers. As co-founder I led product and go-to-market: the survey and interview research, the design of the whole beta, and the pitch deck.
The team won the pitch competition at Johns Hopkins's annual business plan competition, took first place and funding at the Johns Hopkins Spark Accelerator, and joined the NSF I-Corps program. We had three advisors: the head of a computer-oriented learning lab at Johns Hopkins, a managing partner at an education consulting firm that advises companies with $50 million to $2 billion in revenue, and a former business-strategy lead at Ngrade (acquired by McGraw Hill for $55 million) who also served as a special assistant to President Bush at the U.S. Department of Education.
Problem
Teachers spend hours every week building lesson plans, time that comes out of teaching, feedback, and their own lives. General-purpose AI could help but wasn't shaped for how teachers plan, find materials, or keep files.
North starHours a teacher spends planning each week.
Decisions
Research before designing
- Evidence
- More than 200 survey responses and two rounds of user interviews with teachers.
- Chose
- Design the beta around what teachers said they needed: a conversational AI for planning, resource search, file management, and dashboards.
- Didn't build
- Features we couldn't tie back to a survey or interview finding.
- Cost
- Research took time we could have spent shipping.
Put a working beta in real classrooms early
- Evidence
- Teachers in five schools used the beta.
- Chose
- Get the product in front of teachers in five schools and measure planning time there, rather than keep polishing before release.
- Cost
- The beta was rough, and it made our gaps visible to the people we most needed to impress.
Close it down
- Evidence
- The co-founders' priorities had diverged.
- Chose
- Shut down the company instead of continuing with a team that wasn't pulling in one direction.
- Cost
- Ending a product with traction, funding, and five schools behind it.
Solution
The beta gave teachers a conversational AI to plan lessons, a way to search resources, file management, and dashboards. I designed every screen; wireframes are shown below.
Outcome
Teachers using the beta cut lesson planning time by 75%, getting about nine hours a week back. Five schools used it. We won the pitch competition at the Johns Hopkins business plan competition and first place with funding at the Spark Accelerator.
We closed the company in February 2024 because the co-founders' priorities had diverged.
Do differently
- Agree on the commitment each founder was making, in writing, before taking funding and competition prizes.
- Get planning-time measurement into the product from the first beta, so the 75% figure is a tracked number and not a survey answer.
- Line up an engineering owner before the pitches, so the plan didn't depend on one co-founder's availability.