Project-Based Learning That Scales

AI that amplifies coaches instead of replacing them. Stronger teams with real-time project guidance. Learning that develops critical thinkers, not AI dependents.

Actively in development
Faculty-driven project

Project-based learning works — but it's hard to manage and impossible to scale.

We're exploring whether AI can help—not by doing the teaching, but by extending when and where instructors can be present.

This isn't about automation. It's about asking: what if teams had a thinking partner available at 10pm when they're stuck?

What We're Building Toward

These capabilities represent our design vision. We're actively developing and testing these ideas with faculty collaborators.

Team Project Coaching

In Development

Guidance when you can't be in the room

What if teams had coaching available when you can't be? We're exploring how AI might help teams work through the messy parts—defining tasks, setting priorities, staying focused—without waiting for the next class session.

  • Help teams write clear, actionable tasks
  • Surface vague user stories and fuzzy priorities
  • Keep work sessions focused and productive
TeachPilot user story feedback interface
TeachPilot reflection conversation interface

Reflection in the Moment

In Preview

Turn experience into lasting learning

End-of-project reflections are too late—learners have forgotten their reasoning and lost the chance to actually learn. We're designing prompts that surface reflection while decisions are fresh, when insights can still shape what happens next.

  • Ask the right questions while thinking is still accessible
  • Surface assumptions before they become repeated mistakes
  • Build habits that compound across every future project
“Your reflections show growing comfort with ambiguity—a key professional skill.”
TeachPilot skills tracking interface

Support That Knows When to Step Back

In Development

Building independence, not dependency

Most AI tools give the same generic help to every learner, creating reliance instead of growth. We're designing a system where instructors curate the resources and guidance, and the AI calibrates support to where each learner actually is—then fades as competence builds.

  • Deliver instructor-curated resources exactly when they're relevant
  • Adjust scaffolding based on demonstrated skill and reflection depth
  • Reduce support as mastery increases so independence becomes the goal

Your teaching amplified, not replaced.

Instructor Visibility & Insight

In Development

See the learning, not just the final output

We're building toward real-time insight into:

  • Learner thinking paths and decisions
  • Iterations and learning patterns
  • Where intervention has biggest impact

The goal: precise coaching and fairer assessment—without increasing workload.

TeachPilot learner journey timeline

Authentic Project Environment

In Development

Real-world challenges with meaningful structure

We're creating an environment designed for:

  • Open-ended problem solving
  • Collaboration and iteration
  • Realistic constraints and stakeholders

Projects should feel like practice—not just assignments.

Hands-on project work with electronics

Our Design Principles

These beliefs guide what we're building. We'd love to know if they resonate with your experience.

PBL should develop reflective thinkers, not just task performers

Deep learning can be feasible at scale—if we rethink support

The process matters as much as the output—make thinking visible

AI should amplify instructor presence, not replace it

Frequently Asked Questions

Learn more about Project-Based Learning and the vision behind TeachPilot.

Help Shape What This Becomes

TeachPilot is an active experiment. If you're wrestling with the same PBL challenges, we'd love your perspective—whether that's trying early features, sharing your course context, or just telling us what you'd actually need.