explain iq

learn faster by teaching Ai what you know explain your concept in your own words and ai shows you exactly what your missing

Comment

Inspiration

ever finished studying, and felt totally ready, then froze when someone asked you to explain the topic out loud? this question inspired me to build this project, i understood that its not a memory problem but illusion of understanding, recognising answer is far easier than producing one, most learning apps never catch that gap because they only ask you to select an answer, not explain it. i built explainiq to close that gap directly

What it does

explain iq flips traditional studying methods, instead of reading and picking multiple choice answers, a student explains a concept in their own words, ai evaluates the explanation, gives an understanding scor e out of 100, and show exactly what the student got right and whats missing so they can improve and try again, the student becomes the teacher

How we built it

i built with Next.js, (App router) , typescript, and tailwind css on the front end, the backend uses a Next.js route that sends the students explanation to groqs llm api for real time evaluation and structured feedback deployed this MVP on vercel. the design draws inspiration from Duolingo's friendly, engaging learning experience

Challenges we ran into

i initially tried openai's api but hit insufficient quota with no credits available, i pivoted to googles gemini's api, which connected successfully but also hit quota limits during testing, so i switched to groq which offered a generous free tier and fast response time , i also hit front and backend mismatch where the explain from sent (text) but the api route expected (explanation), causing a runtime crash that took focused debugging to trace a single field name

Accomplishments that we're proud of

shipping a fully working ai feedback loop explanation in, structured evaluation out live and deployed under real time pressure with a ui that doesnt feel like a generic hackathon template.

What we learned

how much a single mismatched field name between front end and backend can break an entire feature, and how valuable it is to have a hardcoded fallback response ready so the demo never depends entirely on external api being unavailable

What's next for explain iq

gamification (XP, streaks, levels) inspired by duolingo, a student dashboard tracking weak topics over time, an ai tutor mode that asks socratic follow up questions instead of just scoring, and a classroom mode so teachers can monitor understanding across an entire class

Built With

Share this project:

Updates