PulsePilot

Inspiration

Focus tools usually ask people to work the same way every day. PulsePilot takes the opposite approach: the next task should adapt to the person's current energy and attention state.

What I built

PulsePilot is a lightweight focus companion. It lets a user start a focus session, add a task, and receive an adaptive next-step recommendation. The live prototype stores the user's focus profile and task state through the Evorozen Neural Pulse API, then presents that state in a simple browser interface.

How it works

1. The user starts a session and records a current energy signal.
2. PulsePilot sends the profile and task state to Neural Pulse.
3. The app chooses a practical next action: begin, break the task down, or take a short reset.
4. The updated state is shown immediately so the user can keep moving without planning overhead.

The current demo uses the documented deterministic Neural Pulse profile and task actions so the flow remains reproducible for judges. It does not claim production users or measured traction.

What I learned

The most important lesson was to keep the integration boundary explicit. A small, well-defined profile/task schema made it possible to validate the API independently before building the UI around it.

Challenges

The natural-language provider path was not reliable during the build, so I designed the demo around the documented deterministic API actions instead of hiding that limitation. I also kept the private integration key out of the public repository while still providing a judge-accessible demo.

What's next

Future versions can add richer signals, calendar-aware planning, accessible keyboard flows, and a server-side proxy with stronger authentication and observability.

Built With

• css
• evorozen-neural-pulse-api
• github
• html
• javascript
• rest-api