ChalkTalk: talks-and-draws AI tutor (demo video + live link)

README.md

@@ -1,13 +1,36 @@
 ---
-title: Chalktalk
-emoji: 🐢
+title: ChalkTalk
+emoji: 🎓
 colorFrom: indigo
-colorTo: red
+colorTo: purple
 sdk: gradio
-sdk_version: 6.18.0
-python_version: '3.13'
 app_file: app.py
 pinned: false
+short_description: An AI tutor that talks and draws, live.
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🎓 ChalkTalk
+
+**The AI tutor that talks *and* draws — live, at the same time.**
+
+You type a topic and a tutor explains it out loud while a whiteboard fills in beside it. Every labeled part lights up the exact moment the tutor says its name.
+
+## Watch the demo
+
+<video controls src="https://huggingface.co/datasets/squaredcuber/chalktalk-demo/resolve/main/chalktalk-demo.mp4"></video>
+
+Direct link: [https://huggingface.co/datasets/squaredcuber/chalktalk-demo/resolve/main/chalktalk-demo.mp4](https://huggingface.co/datasets/squaredcuber/chalktalk-demo/resolve/main/chalktalk-demo.mp4)
+
+▶ **Live demo:** [https://belong-picking-platinum-represents.trycloudflare.com](https://belong-picking-platinum-represents.trycloudflare.com)
+
+## How it works
+
+- **PersonaPlex** — a full-duplex, Moshi-architecture 7B speech model that listens and speaks at the same time, so you can interrupt and ask questions mid-explanation.
+- **Nemotron SVG director** — an NVIDIA Nemotron model that draws a clean, labeled SVG diagram for each part of the lesson, streamed to a live whiteboard.
+- **Synced reveal** — the tutor inner-monologue drives the board: each labeled element appears the instant its keyword is spoken, with synonym-aware matching so paraphrases still land.
+- **Auto-progressing lessons** — it teaches one part deeply, then voice and visuals advance together, over one persistent connection.
+- **Runs locally on a single GPU**, wrapped in a shader-animated Gradio app.
+
+## Credits
+
+Built by **Codex**, with help from other coding agents using **GPT-5.5** when parallelization was needed.

app.py

@@ -0,0 +1,19 @@
+import gradio as gr
+VIDEO_URL = "https://huggingface.co/datasets/squaredcuber/chalktalk-demo/resolve/main/chalktalk-demo.mp4"
+LIVE_DEMO = "https://belong-picking-platinum-represents.trycloudflare.com"
+INTRO = """# 🎓 ChalkTalk
+### The AI tutor that talks **and** draws — live.
+Type a topic and a full-duplex voice tutor (**PersonaPlex**) teaches it out loud while an **NVIDIA Nemotron** director draws a labeled diagram on a whiteboard. Every part lights up the moment the tutor says its name.
+"""
+HOW = """
+**How it works** — PersonaPlex (Moshi-arch 7B) speaks and listens simultaneously; a small Nemotron SVG director draws each part; the inner-monologue text reveals each labeled element as its keyword is spoken (synonym-aware); lessons auto-progress with voice and visuals in sync. Runs locally on one GPU.
+
+_Built by Codex, with help from other coding agents using GPT-5.5 when parallelization was needed._
+"""
+with gr.Blocks(theme=gr.themes.Soft(primary_hue="indigo"), title="ChalkTalk") as demo:
+    gr.Markdown(INTRO)
+    gr.Video(VIDEO_URL, label="Demo — a real ChalkTalk session", autoplay=False)
+    gr.Markdown(f"### ▶ Try the live demo: [{LIVE_DEMO}]({LIVE_DEMO})")
+    gr.Markdown(HOW)
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1 @@
+gradio