ORBITAL-X

An AI-powered Digital Bid Manager that helps minority-owned small businesses compete for complex government contracts by automating discovery, compliance, and proposal writing.

Comment

🚀 Inspiration

Government contracts are a $600 billion opportunity in the US alone, intended to uplift small businesses and minority-owned enterprises (MBEs). However, the reality is starkly different. The application process is a maze of 50-page PDFs, complex compliance jargon, and rigid technical requirements.

Small business owners—often single founders or family-run operations—simply do not have the resources to hire "Bid Management Teams" or "Sales Engineers." They see a complex RFP and give up, leaving the field open for giant corporations.

We built ORBITAL-X to level the playing field. We wanted to give every small business owner an expert "Bid Team" in their pocket, powered by Agentic AI.

💡 What it does

ORBITAL-X is an autonomous agentic system that acts as a full-cycle RFP response team. It doesn't just "write" proposals; it thinks through them.

  1. Sales Agent (The Scout): Scans government portals and filters opportunities based on the business's specific capabilities, rejecting anything with unrealistic deadlines (e.g., < 3 months away).
  2. Technical Agent (The Expert): Reads complex technical requirements (e.g., "ISO 9001 certified construction materials") and automatically matches them to the small business's existing inventory or services, explaining why they are a fit.
  3. Pricing Agent (The CFO): Prevents the #1 killer of small businesses: Underbidding. It detects hidden costs like "Mandatory Lab Testing" or "Staggered Delivery Fees" and automatically adds them to the quote, protecting the owner's profit margins.
  4. Integrity Guardrail: Before generating the final bid, our unique "Integrity Index" audits the proposal. If the match is weak or the requirements are vague, it flags the bid for review, preventing accidental non-compliance.

⚙️ How we built it

  • Frontend: React.js + Tailwind CSS for a high-contrast, accessible dashboard.
  • AI Brain: Google Gemini 1.5 Pro via the Generative AI SDK for reasoning and context understanding.
  • Architecture: A client-side "Agent Swarm" where the output of one agent (e.g., the filtered RFP) becomes the input for the next (the matched inventory).
  • Logic: Custom TypeScript logic for the Integrity Index, a weighted algorithm that serves as a safety valve against AI hallucinations:

$$ Integrity = 0.4(SpecMatch) + 0.3(PricingAccuracy) + 0.3(AIConfidence) $$

🚧 Challenges we ran into

  • Hallucinations: Early versions of the AI were "too nice"—they would match any product to any requirement just to be helpful. We had to engineer a "Skeptical Persona" and hard-code guardrails that penalize vague inputs, ensuring the system only recommends bids it is confident in.
  • Complex Pricing: Government RFPs often hide costs in the fine print. We taught our Pricing Agent to hunt for specific keywords like "Warranty," "Certification," and "Logistics" to ensure these costs were captured in the final price.

🏆 Accomplishments that we're proud of

  • The Integrity Index: Building a visual "Trust Score" (Red/Green gauge) that instantly tells a non-technical business owner if a bid is safe to submit.
  • Zero-Latency Interface: By running the orchestration client-side, we achieved a "real-time" feel, making the tool feel responsive and empowering.

📖 What we learned

We learned that access to information is not the problem; access to processing power is. Small businesses can find the RFPs, but they can't process them fast enough. AI isn't just for efficiency; it's an accessibility tool that can bridge the gap between "Main Street" and "Federal Contracting."

🔮 What's next for ORBITAL-X

  • Integration: Connecting directly to SAM.gov (System for Award Management) for live contract feeds.
  • Multilingual Support: Offering the interface in Spanish and other languages to support immigrant founders.
  • Micro-Loans: Integrating with fintech partners to offer working capital loans based on the "Integrity Score" of a drafted proposal.

Built With

Share this project:

Updates