StructPact

StructPact: AI-Powered Construction Negotiation & Payment Protection

Problem Statement #30: AI Agent Spend Controls Track: Blockchain & Crypto

Inspiration

Our journey began with a simple question: What if AI could help businesses negotiate better financial agreements while ensuring that subsequent transactions followed their approved terms?

Initially, we envisioned a general B2B negotiation platform. However, we realized that focusing on one industry would allow us to solve a more specific and meaningful problem.

That's when we identified construction.

General contractors and subcontractors depend on one another, yet their financial priorities often conflict. General contractors want to preserve cash flow and minimize upfront payments, while subcontractors need sufficient funding to purchase materials, pay workers, and keep projects moving.

These competing priorities can lead to prolonged negotiations, cash-flow shortages, payment disputes, and costly delays.

We recognized an opportunity to address both sides of the problem: helping construction businesses reach financially workable agreements while ensuring that subsequent payment requests comply with those agreements.

That's how StructPact was born.

What It Does

StructPact is an AI-assisted construction negotiation and payment protection platform that helps general contractors and subcontractors reach financially feasible agreements and verify subsequent payment requests.

Rather than treating negotiation and payment protection as separate processes, StructPact connects them through six stages:

Project Setup: Enter contract values, construction milestones, projected expenses, and financial constraints.
Negotiation: Use Gemini-powered negotiation or a deterministic solver to evaluate payment schedules that balance both parties' priorities.
Financial Analysis: Compare proposed payment schedules, projected cash flow, and funding shortfalls to assess financial feasibility.
Deal Passport: Convert negotiated terms into a version-controlled agreement with a SHA-256 fingerprint and required approval from both parties.
Payment Protection: Verify simulated payment requests against seven safeguards, including approval status, agreement version, recipient, amount, milestone authorization, contract cap, and duplicate protection.
Audit Trail: Record financial decisions and provide transparent explanations for approved or blocked payment requests.

StructPact's key innovation is connecting negotiation directly to payment verification. The financial terms both parties approve become the rules used to evaluate subsequent payment requests.

How We Built It

We built StructPact as a web-based prototype using OpenAI Codex connected to our GitHub repository, allowing us to develop, test, and refine the application throughout the hackathon.

We integrated Google's Gemini API to provide AI-powered negotiation alongside a deterministic solver. Both approaches evaluate potential milestone payment schedules against the financial priorities and constraints of general contractors and subcontractors.

To ensure financial accuracy, we implemented server-side validation and cash-flow calculations that evaluate projected expenses, incoming payments, and maximum funding shortfalls. AI-generated proposals must satisfy the same financial rules as deterministic proposals.

Once an agreement is selected, StructPact generates a version-controlled Deal Passport with a SHA-256 fingerprint. Both parties must approve the agreement, and amendments require renewed approval.

We connected these agreements to an independent payment firewall that verifies simulated payment requests against seven safeguards. Requests that violate approved terms are rejected and recorded.

We implemented persistent records of agreements, simulated transactions, and audit history to support transparency and accountability.

For our promotional video, we used Higgsfield and Claude to combine AI-generated storytelling with our product demonstration.

Our focus was building an integrated financial workflow rather than disconnected features. Payment execution is simulated for this hackathon; no real financial transactions occur.

Challenges We Faced

One of our biggest challenges was balancing the competing financial needs of general contractors and subcontractors.

A payment schedule that preserves cash for a general contractor may create a significant funding shortage for a subcontractor. We needed to evaluate both parties' constraints while identifying financially feasible agreements.

Financial accuracy was another major challenge. Rather than relying entirely on AI-generated recommendations, we implemented deterministic calculations and server-side validation to verify proposed payment schedules.

We also needed to ensure that approved agreements could not be silently modified or bypassed within our simulated payment workflow. This led us to develop version-controlled Deal Passports, dual-party approval, and independent payment safeguards.

Finally, integrating negotiation, financial analysis, agreement approval, and payment verification into one functional application within a limited timeframe required continuous testing and refinement.

Accomplishments We're Proud Of

Our biggest accomplishment was building a functional prototype that connects AI-assisted negotiation with financial analysis, agreement approval, and payment protection.

We demonstrated two measurable outcomes:

Reduced Projected Funding Shortfall by 16.7%

In our demonstration, StructPact identified a revised payment schedule that reduced a subcontractor's maximum projected funding shortfall from $18,000 to $15,000, a $3,000 improvement.

This demonstrated how adjusting milestone payment timing can improve projected liquidity without necessarily changing the total contract value.

Blocked a Duplicate $17,500 Payment

Our payment firewall detected and rejected a duplicate $17,500 milestone payment request within the simulated ledger.

By comparing the request against approved agreement terms and existing records, StructPact demonstrated how payment verification can help prevent unauthorized transactions.

We're also proud of implementing Gemini-powered negotiation, deterministic financial validation, SHA-256 agreement fingerprints, seven payment safeguards, and persistent SQLite audit records.

Most importantly, we transformed an ambitious concept into a working, integrated prototype rather than stopping at an idea or presentation.

What We Learned

Building StructPact taught us that effective AI-powered financial tools require more than simply integrating a language model.

We learned the importance of combining AI-assisted decision-making with verifiable calculations, human oversight, and transparent financial controls. AI can help explore potential agreements, but financial decisions must still satisfy explicit constraints.

We also gained hands-on experience with AI-assisted development, API integration, financial modeling, server-side validation, debugging, and iterative testing.

Our biggest takeaway was that meaningful innovation doesn't always require inventing an entirely new technology. Sometimes, it comes from connecting existing capabilities in a way that solves a problem more completely.

What's Next

Our vision is to develop StructPact into a trusted financial coordination platform for the construction industry.

Our next steps include creating secure, separate accounts for general contractors and subcontractors, expanding AI negotiation capabilities, and improving cash-flow forecasting.

We also plan to integrate with construction accounting and project management software, support verified construction milestones, and accommodate more complex agreements involving multiple subcontractors and suppliers.

Eventually, we envision integrating real payment infrastructure, subject to appropriate security, authorization, and compliance requirements.

Before pursuing live transactions, we aim to validate the platform with construction professionals and better understand their existing workflows.

Ultimately, we want StructPact to help construction businesses spend less time negotiating payment conflicts and more time completing projects.

StructPact: Negotiate smarter. Protect cash flow. Pay securely.