DealMile

AI-powered project agreements that turn client requirements into clear milestones, detect scope changes, and connect approved work to PayPal payments.

Comment

DealMile

Inspiration

Freelance projects often begin with unclear requirements, poorly defined deliverables, and disagreements about what was originally agreed upon. As projects evolve, scope changes can create additional confusion between clients and developers.

We wanted to build a platform that makes freelance collaboration more structured, transparent, and predictable.

DealMile was inspired by the idea that better agreements and milestone-based payments can help both parties establish trust before committing to project work.

What it does

DealMile is an AI-powered freelance project management and agreement platform that connects clients and developers through structured proposals, milestone-based funding, and intelligent scope management.

Key features include:

  • AI Project Planning: Generate structured project scopes, milestones, and deliverables from natural-language requirements using Gemini AI.
  • Collaborative Agreements: Allow clients and developers to review, accept, reject, or request revisions to project proposals.
  • Milestone-Based Funding: Integrate PayPal Sandbox for milestone payments with server-side transaction verification.
  • Scope Intelligence: Analyze requested project changes against the original agreement to identify potential scope differences.
  • Additional Milestones: Support proposals for additional work with separate budgets and timelines.
  • Developer Scheduling: Visualize project milestones and schedules through an interactive calendar.
  • Project & Financial Tracking: Monitor project progress, milestone funding, and financial records through dedicated dashboards.

How we built it

DealMile was built using a modern full-stack architecture:

  • Next.js 16 & React 19 for the application interface and server-side functionality.
  • TypeScript for type-safe development.
  • Prisma & PostgreSQL for database management and persistent project data.
  • Google Gemini AI for project planning and scope intelligence, using structured outputs validated with Zod.
  • PayPal Sandbox for payment order creation, approval, capture, and transaction verification.
  • AG Grid for portfolio and financial data visualization.
  • Bryntum Calendar for developer scheduling and milestone timeline visualization.

The application was designed around structured project workflows, explicit agreement states, and server-side verification for critical payment operations.

Challenges we ran into

One of our biggest challenges was designing a workflow that connects AI-generated project planning with actual collaboration between clients and developers.

AI-generated requirements needed to remain editable rather than becoming unquestionable decisions. We also had to distinguish original project scope from additional requests while keeping humans responsible for final approvals.

Another major challenge was integrating PayPal into milestone-based workflows. Payment approval and payment capture are separate steps, so the application needed to verify transaction details on the server before updating funding records.

Managing project state transitions, payment consistency, and additional milestone approvals also required careful consideration.

Accomplishments that we're proud of

  • Built an end-to-end freelance collaboration workflow, from AI-assisted project creation to milestone funding and scope changes.
  • Integrated Gemini AI into both project planning and scope analysis.
  • Implemented PayPal Sandbox payment flows with server-side verification.
  • Created a structured proposal review and acceptance workflow.
  • Developed Scope Intelligence to help identify changes beyond the original agreement.
  • Built interactive portfolio, financial ledger, and developer scheduling interfaces.
  • Established a cohesive product identity and user experience around DealMile.

What we learned

We learned that AI is most useful when it supports human decisions rather than replacing them.

Structured outputs, validation, and editable AI-generated plans are important when AI results directly influence project agreements.

We also learned that payment integrations require more than a successful redirect. Transaction verification, state consistency, and handling the distinction between approval and capture are essential.

Most importantly, trust in collaborative platforms comes from clear workflows, transparent information, and predictable decisions—not simply from adding AI to an application.

What's next for DealMile

DealMile has a foundation for further development. Our next steps include:

  • Moving beyond Sandbox into properly configured production payment workflows.
  • Expanding automated testing across project, agreement, and payment state transitions.
  • Improving project revision workflows and communication between clients and developers.
  • Enhancing AI scope analysis with more detailed change-impact assessments.
  • Developing more comprehensive project analytics and financial reporting.
  • Exploring a broader developer ecosystem and improved onboarding experience.

Our long-term vision is to make DealMile a reliable collaboration platform where clients and developers can agree on work, manage changes, and track milestone-based progress with greater confidence.

Built With

Share this project:

Updates

Submission history