October workshops are openBuild a Search AI Agent$99 early bird

Enroll now
Skip to content

Education · INTERACTIVE AGENT DEMO

Next-Step Learning Coach

Explore how a learning assistant chooses useful next-step feedback.

Change the briefRun working toolsInspect the result

YOUR TURN TO EXPERIMENT

Free to explore · No sign-in or API key

Give the agent a brief.

Work through an original algebra problem across several turns. Enter your next step, inspect the check, and adjust how much help the coach provides.

The learner has correctly undone the addition.

Working local tools. Authored controller and classroom data. No live AI model or external action.

A SMALL BRIEF. A VISIBLE RESULT.

See the agent at work.

Start with a scenario, change a detail, and run the demo. Your result and the steps behind it will appear here.

1 · Read the brief2 · Use tools3 · Check the result

Try this twistFirst, solve a problem through two or more runs. Then revisit an incorrect step at hint levels 1 and 3 and compare how much guidance each gives.

Open the code behind this agentRead it, change it, run your own variation.

This is the JavaScript that powers the demo. agent.js coordinates the workflow; tools.js contains its available tools. Tool calls and returned data are inspectable after each run.

Ready when you are

⌘ / Ctrl + Enter runs your agent.js draft with the current inputs. Tab leaves the editor. Package imports and network calls are unavailable in this bounded environment.

Ready for a real model and open-source tools? Explore the separate local SDK setup

The real research behind this demoWhere the idea comes from, and what this classroom version does.

Google's LearnLM research studies how models follow teaching instructions. This adaptation lets you inspect a simpler teaching decision: selecting feedback from the learner's current attempt and requested hint level.

The equation checker really runs, but only understands the small documented algebra format. It calls no model, uses no grades, and does not assess a learner's abilities.

STEMtiq authored this exercise. The named organizations do not endorse it. The JavaScript tools run locally; the controller is programmed, and classroom records are labeled examples. This is not a live instance or reproduction of the referenced research system.

LearnLM Team · Google

LearnLM: Improving Gemini for Learning

LearnLM studies following pedagogical instructions, including interactive tutoring that stays on topic and avoids giving away answers prematurely.

The paper's expert preference evaluation does not establish this demo's learning outcomes. Our authored hint bank and equation checker are original and do not run LearnLM.

Where will your curiosity go next?

Your JavaScript really runs. The model decisions and school data are authored simulations, so you can learn without an API key. Every workspace also includes a separate real SDK example to explore next. Passing the lab’s cases is practice, not proof that an agent is ready for real-world use.