AEI4U

Analog Electronic Intelligent for You β€” an AI teaching assistant for Tsinghua's Fundamentals of Analog Electronics. πŸ† Grand Prize, 43rd Tsinghua Challenge Cup.

AEI4U (Analog Electronic Intelligent for You) is a course-grade AI teaching assistant for Fundamentals of Analog Electronics, built on the Dify-THU platform. It closes the loop from theory to simulation to personalized practice β€” and turns β€œlow-code AI product landing” into a plug-and-play pattern for university courses.

My role. Project Lead. Award. πŸ† Grand Prize (η‰Ήη­‰ε₯–), 43rd Tsinghua Challenge Cup β€” AI Teaching-Agent Design Track. Platform. Dify-THU. Course. Fundamentals of Analog Electronics (γ€Šζ¨‘ζ‹Ÿη”΅ε­ζŠ€ζœ―εŸΊη‘€γ€‹), Tsinghua.

What it does

AEI4U fuses three pillars into one assistant:

  1. Circuit topology recognition β€” students upload a hand-drawn or screenshotted schematic; AEI4U identifies the circuit and grounds the conversation in it.
  2. Simulation in the loop β€” connected to SPICE for real experimental feedback, not just textbook formulas.
  3. Personalized learning profiles β€” tracks what each student struggles with and shapes the next interaction accordingly.

Pain points it addresses

  • Analog electronics is abstract β€” circuit principles are hard to internalize without simulation.
  • Lab operation is complex β€” students need scaffolded guidance, not just answers.
  • Office hours don’t scale β€” a 24/7 course-grade assistant fills that gap.

Architecture

  • Dify-THU workflow backbone β€” low-code agent definition that’s easy to iterate.
  • API-first front-end / back-end split β€” embeddable into other learning platforms.
  • Knowledge graph β€” covers theoretical content of the course.
  • SPICE bridge β€” experimental simulation called as a tool.
  • User profile module β€” drives adaptive interaction.

Why this template generalizes

The bigger contribution is AEI4U-as-a-pattern: a low-code template that other courses can drop in. We treat course-AI as a product-engineering problem β€” one that any STEM department can adopt without bespoke ML.

  • πŸ›οΈ Tsinghua’s 43rd Challenge Cup β€” AI Teaching-Agent Design Track (call announcement)
  • πŸ“„ Award materials (poster, full report, demo video) live in the project’s local archive β€” drop into assets/img/ to display here.
Cover image is a placeholder. Export slide 1 of AEI4Uζ΅·ζŠ₯.pptx as PNG and replace assets/img/aei4u_cover.svg.