Proposeer - Bid smarter, not harder

💡 Inspiration Proposeer was born from a real-world bottleneck in the high-stakes world of Federal Contracting. While consulting with a partner at a Washington Beltway firm, we identified a critical inefficiency: the "RFP Rabbit Hole."

The Problem: Teams were spending weeks manually parsing massive documents, coordinating Subject Matter Experts (SMEs), and cross-referencing past performance.

The Risk: Often, they realized a bid wasn't financially or strategically viable only after wasting hundreds of billable hours.

The Vision: We built Proposeer to transform this chaotic manual process into a streamlined, data-driven decision engine.

🚀 Challenges & Evolution The Legacy Gap The vision for Proposeer actually began six years ago, but we were limited by the technical constraints of the era. Early iterations relied on traditional Machine Learning to parse thousands of pages and correlate bid data with win/loss outcomes. However, mapping complex variables—like SME availability and the strategic value of joint ventures—remained a massive technical hurdle.

The Gemini Breakthrough The true breakthrough came with the integration of Gemini AI. By leveraging the Gemini 3 Flash model, we transitioned from basic text parsing to sophisticated semantic understanding.

Key Milestone: Despite the initial learning curve and the hurdles of managing API rate limits during development, Gemini’s ability to process massive PDF datasets allowed us to finally realize the project’s full potential.

The Result Proposeer is now a predictive engine that understands more than just the text on the page; it understands the strategic context of how to win the contract.

Built With

• gemini
• node.js
• typescript