Market Opportunity Scout

A "Super Great" automated engine for tech firms. Use Gemini’s reasoning to analyze complex project portals, manage persistent lead sources, and get real-time alerts on urgent global bids.

Comment

Inspiration

As a software developer managing a remote team, I realized that the biggest barrier to growth isn't technical skill—it’s information symmetry. High-value IT opportunities are scattered across hundreds of global portals, each with complex requirements. I was inspired to build Market Opportunity Scout to eliminate the manual "hunt" and replace it with an intelligent, automated engine that allows African tech firms to compete on a global stage.

What it does

Market Opportunity Scout is an AI-driven command center for tech entrepreneurs.

Multi-Link Intelligence: It processes up to 3 different project URLs in parallel, using Gemini to extract core requirements.

Smart Persistence: A built-in Sources Database tracks the most fertile lead environments.

Subscription Engine: Users can "subscribe" to specific portals, triggering automated scans for new "popped up" bids.

Real-time Awareness: A sleek notification system and dashboard track "Urgent Bids" and active sources in real-time.

How we built it

The platform is built with a modern, high-performance stack designed for scale:

Frontend: Next.js 15+ and TypeScript for a type-safe, responsive UI.

AI Engine: Google Gemini 2.0 Flash via the API, utilizing its high context window to reason through multiple tender documents simultaneously.

Backend: FastAPI provides a high-speed asynchronous bridge between the AI and the database.

Storage: SQLite ensures persistent, lightweight tracking of the top 10 most-used sources.

Styling: Tailwind CSS and a custom "Toast" notification system for a premium UX.

Challenges we ran into

One major challenge was Schema Extraction. Different tender portals use vastly different HTML structures. We overcame this by moving away from traditional scraping and instead passing the raw structured data to Gemini with specific prompts to "reason" and normalize the output into a unified JSON format. We also had to optimize the parallel processing of three concurrent links to ensure the UI remained responsive during heavy AI analysis.

Accomplishments that we're proud of

Parallel Reasoning: Successfully implementing the "3-Box" analysis modal where the AI handles multiple contexts without losing accuracy.

Persistent Intelligence: Building a system that actually remembers which sources are most valuable to the user.

User-Centric Design: Creating a dashboard that transforms complex data into actionable "Urgent Bid" stats, making it easy for a CEO or Project Manager to make decisions at a glance.

What we learned

During this build, we deepened our understanding of prompt engineering for data normalization. We learned that Gemini is exceptionally good at identifying "Urgency" in business text—a nuance that traditional algorithms miss. We also learned the importance of "Building in Public" and how real-time feedback shapes a project from a simple tool into a "Super Great" platform.

What's next for Market Opportunity Scout

The next step is Multimodal Tender Analysis. We plan to enable Gemini to analyze PDF blueprints and complex architectural attachments often found in large-scale IT bids. We also aim to integrate a "Team Assignment" feature, allowing the scout to automatically suggest which member of a remote team is the best fit for a specific opportunity based on their skills.

Built With

Share this project:

Updates