PactFlow — From Commitments to Action

Inspiration

Every day, we make commitments: preparing for exams, completing projects, meeting deadlines, learning new skills, and working toward long-term goals.

We have calendars to manage our time, task managers to organize responsibilities, and AI assistants to answer questions. Yet these tools often operate independently.

They help us remember what we need to do, but they don't necessarily tell us when our plans have become impossible.

As a student balancing academics, programming, personal projects, and deadlines, I wanted to explore a more proactive kind of personal AI.

What if an assistant could remember my priorities, understand my available time, identify conflicts before they become problems, and help me make realistic decisions?

That question inspired PactFlow.

What It Does

PactFlow is a privacy-conscious personal AI assistant designed to bridge the gap between planning and execution.

Instead of functioning only as a conversational chatbot, PactFlow focuses on understanding commitments and helping users take meaningful action.

Its central component is Commitment Guardian, a system designed to examine deadlines, priorities, tasks, and calendar availability to identify scheduling conflicts.

The project is organized around five capabilities:

1. Persistent Personal Memory

PactFlow is designed to retain useful context such as goals, preferences, projects, and commitments across sessions.

This allows recommendations to consider the user's priorities rather than treating every interaction as an isolated conversation.

2. Commitment Guardian

The flagship feature focuses on detecting conflicts between deadlines and available time.

For example, if a student has an examination tomorrow but has allocated most of the evening to a flexible project, PactFlow can identify the competing commitments and suggest a more realistic arrangement.

3. Intelligent Scheduling

PactFlow combines task priorities, estimated durations, and scheduling constraints to generate practical planning recommendations.

The goal is to help users understand trade-offs rather than simply fill every available hour.

4. Human-Controlled Automation

Autonomy is useful only when users remain in control.

PactFlow's architecture separates proposed actions from their execution, allowing important schedule changes to be reviewed and approved.

5. Explainable Decisions

Instead of producing unexplained recommendations, PactFlow aims to show which deadlines, constraints, and preferences

Built With

• apscheduler
• docker
• fastapi
• javascript
• pydantic
• python
• react
• rest-api
• sqlalchemy
• sqlite
• tailwind-css
• typescript