• 

  Home Page

• 

  How It Works

• 

  Judge Mode ( Example Demo )

• 

  Demo

ShareCircle — Turning Local Needs Into Real-World Help

Inspiration

We kept coming back to a simple question:

What if the person who can help you is already nearby, but you simply don't know they exist?

Communities are full of people with spare time, skills, resources and willingness to help. At the same time, other people nearby may need tutoring, meals, transportation, errands, or small everyday assistance. The problem isn't always a lack of resources — it is often a lack of discovery, trust and coordination.

That idea became ShareCircle: an open-source platform designed to connect local needs with people who can fulfill them.

What We Built

ShareCircle allows users to post either a Need or an Offer.

A user can describe what they need, specify their category, urgency and location, and discover relevant people nearby. Helpers can publish what they are willing to offer.

The platform then guides the interaction through a structured workflow:

Post → Match → Accept → Arrive → Pay → Verify → Complete

We also built OTP verification, messaging, location-based discovery, match tracking, impact receipts and explainable matching information.

For this project, we are extending that workflow with an AI-assisted matching agent. Instead of forcing users to understand filters and categories, they can describe their request naturally. The agent extracts useful information, works with the deterministic matching system to find suitable candidates, and explains why a particular person is a relevant match.

For example:

"I need someone nearby who can pick up medicines for my grandmother tomorrow afternoon."

ShareCircle can turn that into structured information such as the category, urgency, timeframe and location requirements, then surface suitable offers.

Why AI?

We didn't want to add AI just to put an "AI" label on the project.

The useful role for AI here is understanding intent.

Traditional forms require users to know exactly which category, filters and options to select. Real people don't naturally communicate that way. They describe problems.

Our AI layer helps translate those natural descriptions into actionable matching requirements while leaving important constraints such as location and availability to the deterministic matching system.

This gives us a combination of:

• AI for understanding and recommendations
• Structured logic for reliable matching
• Human users for the final decision

What We Learned

Building ShareCircle taught us that a useful product is much more than its central algorithm.

We had to think about what happens before and after a match: location permissions, accessibility, OTP verification, payment confirmation, cancellation, messaging, safety, mobile navigation and what happens when there are no suitable matches.

We also learned that explainability matters. If an application recommends another person for something important, users should understand why that recommendation appeared rather than simply being shown an unexplained score.

Challenges

One of our biggest challenges was designing a matching experience that was useful without becoming complicated.

Location was another challenge. Automatically requesting a user's location can be intrusive, so we changed the experience to make location sharing explicit and provide manual alternatives.

We also had to design the match lifecycle carefully because the platform deals with real interactions between people. OTP verification, payment confirmation, status updates and safety guidance all became important parts of the experience.

Finally, we had to balance AI flexibility with deterministic behaviour. We don't want an AI model deciding everything. Instead, AI helps interpret and recommend while the application's core rules remain predictable and testable.

Open Source

ShareCircle is designed as an open-source project because mutual aid benefits from transparency and community contribution.

The goal is not to build another closed platform where communities depend entirely on a single company. We want organizations, student communities, residential communities and local groups to be able to inspect, adapt and potentially deploy the system for their own needs.

What's Next

The current platform establishes the core mutual-aid workflow. Future development can expand the AI matching agent, improve multilingual support, strengthen privacy-preserving impact records, add richer community analytics and make the platform easier to deploy for different communities.

Ultimately, ShareCircle is built around a simple belief:

Help already exists in communities. We want to make it easier to find.

Built With

• agentic
• ai
• api
• css
• design
• flask
• gps
• html
• javascript
• open
• python
• responsive
• rest
• source
• sqlite
• web