AIZOYA Opportunity Intelligence

A WebMCP-native command center enabling people and AI agents to securely discover, compare, prioritize, and shortlist business opportunities using structured web tools.

Comment

Inspiration

Business owners and innovation teams face more opportunities than they can manually assess. Key information is often scattered across dashboards, websites, documents, and portals. Traditional browser agents must interpret visual interfaces and guess which controls to use, resulting in slow and unreliable performance.

AIZOYA Opportunity Intelligence introduces a new approach: a website designed for both people and AI agents. WebMCP enables the site to present clear, structured capabilities while maintaining the visual experience and user control.

What it does

AIZOYA Opportunity Intelligence serves as a WebMCP-enabled opportunity command center. Users can instruct an AI agent to:

  • List the opportunities available on the page
  • Compare two opportunities using their scores, strategic fit, deadlines, and other decision factors
  • Add a selected opportunity to the shortlist

The agent uses registered website tools rather than guessing through the interface. Because the page and agent share the same context, results appear directly in the user interface.

The prototype shows how WebMCP can support business research, innovation portfolios, grant discovery, accelerator applications, procurement opportunities, and other decision-intensive workflows.

How we built it

We developed a responsive web application with Next.js, React, and TypeScript, and deployed it on Vercel. The application registers structured WebMCP tools with defined names, descriptions, inputs, and outputs.

The core tools support three stages of the decision workflow:

  1. Opportunity discovery
  2. Side-by-side comparison
  3. User-directed shortlisting

The interface displays structured opportunity records, and the WebMCP layer provides agents with reliable access to these capabilities. We tested the deployed application using ChatGPT’s in-app browser by issuing natural-language requests and confirming the correct tools were identified and executed.

Challenges we ran into

Our main challenge was designing tools that were useful to agents without being too broad or ambiguous. Tool descriptions, parameter schemas, opportunity identifiers, and results needed to be predictable.

We also needed to make agent interactions visible and understandable to users. Because shortlisting changes application state, the workflow required clear intent and observable feedback instead of hidden automation.

Testing an emerging browser standard required validating both the traditional user interface and the agent-driven experience.

Accomplishments that we’re proud of

  • Built and deployed a working WebMCP-enabled application
  • Registered multiple focused tools instead of one unrestricted tool
  • Demonstrated an end-to-end agent workflow
  • Enabled structured comparison between opportunities
  • Connected an agent action to visible shortlist state
  • Preserved a useful interface for people who interact with the site manually

What we learned

WebMCP changes how websites communicate their capabilities. Visual design remains important for users, while structured tools give agents clearer instructions and more reliable actions.

We also learned that the most effective agent tools are focused, well-described, and based on visible application state. Human control and transparency become increasingly important as agent capabilities grow.

What’s next

Next, we plan to expand the prototype into a comprehensive AIZOYA portfolio command center with authenticated workspaces, live opportunity sources, evidence-backed scoring, deadlines, reminders, bilingual support, audit history, and approval controls for key actions.

Our long-term vision is a human-centered intelligence system that enables organizations and their AI agents to discover opportunities, evaluate them consistently, and act more quickly and with greater accountability.

Built With

  • ai
  • and-javascript-and-deployed-it-on-vercel.-the-application-uses-the-imperative-webmcp-api-through-document.modelcontext.registertool(.)-to-register-six-structured-tools-with-defined-descriptions
  • annotations
  • chatgpt
  • codex
  • css
  • css3
  • github
  • html5
  • input-schemas
  • javascript
  • validation
  • vercel
  • webmcp
Share this project:

Updates

posted an update —

We found building AIZOYA Opportunity Intelligence so valuable that we chose to develop it beyond a hackathon demonstration.

We are now adapting its human-and-agent collaboration model for AIZOYA’s Project Command Center.

The original prototype demonstrated how WebMCP enables AI agents to list opportunities, retrieve details, compare priorities, identify deadline conflicts, build action plans, and update a shared shortlist using structured website tools.

This approach will support management of AIZOYA’s expanding project portfolio. Our next phase will apply these principles to:

  • View projects and workstreams in one place
  • Compare priorities, risks, deadlines, and strategic value
  • Identify conflicts and missing dependencies
  • Build recommended action plans
  • Maintain a shared, human-approved priority list
  • Give AI agents structured tools instead of making them guess their way through the interface

The hackathon demonstrated that WebMCP is more than a demonstration technology. It can serve as a practical operating layer for coordinating people, projects, and AI agents at AIZOYA.

The prototype developed for the WebMCP Challenge is now shaping the design of our internal command center. This may be the project's most valuable outcome.

— Franklyn Rosario, MBA-ITM Founder & CEO, AIZOYA

Log in or sign up for Devpost to join the conversation.

Submission history