Inspiration
We built LifePilot around a simple problem: people often know what they want to accomplish, but turning that goal into a realistic sequence of actions can be difficult.
For example, saying "I have a software engineering interview tomorrow" is easy. Deciding what to study, what to prioritize, how much time to spend on each topic, and what to do next is much harder.
LifePilot was designed to bridge that gap — from conversation to action.
What it does
LifePilot is a conversational productivity assistant that turns natural-language goals into actionable workflows.
A user can describe a goal such as:
"I have an interview tomorrow. Help me prepare."
LifePilot generates a structured plan containing prioritized tasks, estimated durations, subtasks, and clear next actions.
It can also adapt the workflow when constraints change. For example, when the user says:
"I only have two hours today."
LifePilot creates a 120-minute emergency plan focused on high-yield activities.
The application includes:
Conversational goal and intent handling
Personalized action plans
Priority and time-based task organization
Time-constrained plan adaptation
Task completion and progress tracking
Focus/Pomodoro-style timer
Interactive technical mock interview with feedback
Browser-based voice interaction
Alexa+ inspired simulation
How we built it
LifePilot is a React and Vite web application built with JavaScript.
The application uses client-side intent detection to identify requests such as interview preparation, time optimization, daily planning, next actions, learning Java, mock interviews, and task creation.
Workflow plans and progress are stored locally using browser localStorage, allowing the application to maintain state without requiring a backend database.
The mock interview uses a predefined technical question set with heuristic evaluation to provide immediate feedback.
The voice experience uses browser Web Speech APIs for speech recognition and speech synthesis.
For the Alexa+ track, we implemented the simulated Alexa+ experience path using a browser-based voice interface. This is clearly presented as a simulation rather than a live Alexa+ integration.
Challenges we ran into
One challenge was turning vague natural-language goals into useful, concrete actions without making the user configure a complicated workflow.
Another challenge was adapting plans when the user's available time changed. Instead of simply removing tasks, LifePilot prioritizes high-value activities and restructures the workflow around the available time.
We also wanted the voice experience to remain lightweight and accessible in a browser, which led us to use the browser's built-in speech APIs for the simulation.
Accomplishments that we're proud of
We are proud of creating a working end-to-end productivity experience where a simple conversational request can become a structured workflow.
We are particularly proud of the time-adaptive planning feature, because it turns a common real-world constraint — limited time — into an actionable preparation strategy.
We also integrated task progress, focus sessions, mock interview practice, and a voice-based interaction experience into one cohesive application.
What we learned
We learned that conversational interfaces become much more useful when they lead to concrete actions rather than stopping at recommendations.
Building LifePilot also helped us understand how intent detection, workflow generation, time prioritization, local state management, and voice interaction can work together to create a more natural productivity experience.
Most importantly, we learned to design around the user's immediate goal: what should I do next?
What's next for LifePilot
The next step for LifePilot would be connecting it with external calendars and task-management platforms so generated workflows can become part of a user's existing schedule.
The voice experience could also evolve from the current browser simulation into a production Alexa+ integration using the appropriate Amazon developer technologies.
We also see opportunities to make the planning engine more adaptive by learning from completed tasks, deadlines, user preferences, and changing priorities.
Log in or sign up for Devpost to join the conversation.