Why We Started Impact-Edu.ai

For the past decade, our team has been building AI-powered education tools — adaptive assessments, learning games, computer vision for paper worksheets. These tools have reached 15 million students across the US and India.

But the deeper we got into the work, the clearer a problem became: we don't know if we're teaching the right things.

Education's standards were designed before AI existed. The Common Core has no sense of priority — no way to say which objectives matter most when a student can ask ChatGPT to do their homework. The process of updating standards takes years. Meanwhile, AI is being adopted in classrooms faster than anyone can study its effects.

Well-funded schools experiment. Under-resourced schools fall further behind. Vendors make claims nobody can verify. Teachers make decisions with no evidence to guide them.

Impact-Edu.ai exists to close that gap. We build the open research infrastructure — assessments, datasets, evaluation frameworks, practitioner training — that the field needs to figure out what matters in an AI age and whether we're actually teaching it.

Everything we produce is open: open access, open source, Creative Commons. Because the answers to "what should students learn?" shouldn't be locked behind a paywall.