이기는 전략이,
제안서 끝까지 이어집니다.

The Experience That Led to contrl

contrl was born from direct, firsthand experience with the pain of proposal writing. Our founding team spent years in enterprise B2B sales and government contracting, where responding to RFPs was a recurring, time-consuming process. We watched senior proposal writers spend 2-3 weeks per RFP response — analyzing requirements, developing win themes, writing content, and then spending days formatting everything into the company's PowerPoint template.

When AI tools like ChatGPT emerged, we tried using them for proposal writing. The result was disappointing: the AI could generate text, but it couldn't produce formatted PowerPoint files. We'd spend hours copying AI-generated text into our template, fixing layouts, adjusting fonts, and ensuring brand consistency. The "time saved" by AI was largely consumed by formatting work.

This firsthand frustration led us to build contrl. We analyzed over 500 real RFP documents across government procurement, IT solutions, consulting, and defense sectors to train our system. We reverse-engineered PowerPoint's XML structure to achieve 100% template fidelity. And we built everything with a local-first architecture because we knew that proposal documents contain sensitive competitive intelligence that cannot be uploaded to cloud services.

Technical Expertise Behind contrl

contrl's technology is built on three pillars of expertise:

1. RFP Analysis Engine: Our NLP pipeline is specifically trained on RFP document structures — not general text. It understands the difference between mandatory requirements, evaluation criteria, compliance matrices, and scoring rubrics. It can parse complex RFP formats including government procurement documents (조달청, SAM.gov, GeBIZ), corporate RFPs, and multi-volume solicitations.

2. Template DNA Extraction: We developed proprietary technology to "read" PowerPoint templates at the XML level. This goes far beyond slide layouts — we capture font hierarchies, color schemes, placeholder positions, master slide inheritance, chart styles, table formatting, and even subtle design elements like gradient fills and shadow properties. The result is output that's indistinguishable from manually-created proposals.

3. Win Theme Intelligence: Drawing on our team's experience with hundreds of successful proposals, we built an AI system that doesn't just summarize RFPs — it identifies strategic opportunities. It maps your organization's strengths against the RFP's evaluation criteria to generate actionable win themes that increase your probability of winning (Pwin).

Authority in Proposal Automation

contrl is developed by Cliwant Inc., a company focused exclusively on AI-powered document automation for B2B proposal teams. Our focus is narrow by design — we don't try to be a general-purpose AI tool. This specialization allows us to solve proposal-specific problems that general AI tools cannot address.

Our platform has been used to analyze RFPs and generate proposal drafts across multiple industries including government procurement, defense contracting, IT system integration, management consulting, construction, and telecommunications. We work directly with proposal teams to continuously improve our system based on real-world feedback and win/loss analysis.

contrl is the only AI proposal tool that combines four critical capabilities in one platform: (1) RFP requirement analysis, (2) win theme strategy generation, (3) template-preserving document generation, and (4) source traceability — with your proposal files never leaving your machine.

Trust and Transparency

How contrl generates proposals (process transparency): contrl is not a black box. Every generated proposal includes full source traceability — each paragraph, claim, and data point links back to the specific RFP section, page number, and clause that informed it. Reviewers can verify every piece of content against the original RFP, ensuring accuracy and compliance.

Why your proposal files stay protected (security transparency): We designed contrl so your proposal files never leave your machine — not as a marketing feature, but as a fundamental design principle. Proposal documents contain sensitive competitive intelligence — pricing strategies, technical approaches, personnel qualifications, and proprietary methodologies. Uploading these to cloud AI services creates unacceptable security risks. With contrl, your proposal files remain on your machine, with zero data transmitted externally.

What contrl does NOT do (honest limitations): contrl generates proposal drafts, not final submissions. The AI-generated content requires human review, strategic refinement, and approval before submission. We believe AI should augment proposal teams, not replace them. Our source traceability feature is specifically designed to make human review efficient and thorough.

Data handling policy: contrl never stores, transmits, or uses your documents for model training. Your proposal files — RFPs, templates, and generated proposals — never leave your machine. We do not have access to any document you process through contrl.

Our Development Methodology

contrl's AI engine was developed through an iterative process informed by real proposal workflows:

• Analyzed 500+ RFP documents across 12 industries to build our requirement extraction model
• Studied 200+ winning proposals to understand effective win theme patterns
• Reverse-engineered PowerPoint's OOXML specification to achieve pixel-perfect template reproduction
• Tested with 50+ corporate PPT templates to ensure universal template compatibility
• Conducted blind comparison tests where proposal managers could not distinguish AI-generated drafts from human-written ones

This experience-driven development approach ensures that contrl solves real problems faced by real proposal teams, not hypothetical use cases.