Turning “why is it doing that?” into a diagnosis.
Built an interactive hardware-debugging experience that turns raw device telemetry into guided diagnosis using expected signals, baselines, deterministic checks, and AI-assisted reasoning.
How I built it & what I contributed
- Idea and direction
I developed the idea for ReWeird: a hardware-debugging experience that helps people understand why a device is behaving unexpectedly. I worked with my team to turn that concept into a workflow connecting device signals, diagnostic evidence, and recovery steps.
- Building the experience
I helped build the interactive application around a clear sequence: inspect the device, compare its behavior with an expected baseline, investigate a fault, and try a fix. The project combines a React and JavaScript interface with Go, REST APIs, and CRUD workflows.
- Making telemetry understandable
I worked on turning raw telemetry into an experience people could explore instead of leaving them with unexplained readings. Expected signals and baselines give the diagnostic workflow a reference for identifying abnormal behavior.
- Interactive fault simulation
I developed the direction for the “Make It Weird” demonstration, where visitors introduce a fault and follow what changes. I worked on the judge-facing simulation so people could try the debugging experience with simulated values.
- Evidence and recovery
The diagnostic approach combines deterministic checks with AI-assisted reasoning. I helped shape an experience that connects a suggested diagnosis to evidence and lets visitors test a fix and check whether the device has recovered.
- Demo delivery and collaboration
I worked on preparing the landing page and interactive demo for judges to access through the shared project. Git, GitHub, and GitHub Actions supported a team workflow and delivery of the project.