Skip to main content
The Proposal Writer is the core feature of the Alliance IGAD Innovation Hub, providing a structured 4-step workflow to transform RFP documents into polished proposals using AI assistance.

​
Workflow Overview

The proposal writing process follows a systematic approach:
1

Step 1: Information consolidation

Upload and analyze your RFP document, concept materials, and reference proposals. The AI extracts requirements, evaluation criteria, and key deadlines.
2

Step 2: Concept review

Review AI-generated concept analysis, select sections to include, and provide feedback. The system generates an enhanced concept document aligned with RFP requirements.
3

Step 3: Structure and workplan

Generate a proposal structure based on RFP requirements and your concept. The AI creates a logical outline with sections, subsections, and workplan components.
4

Step 4: Proposal review

Generate a draft proposal template or AI-written draft, then analyze it for completeness, alignment, and quality. Receive actionable feedback for improvement.

​
Step 1: Information consolidation

​
Document upload

Upload the following materials:
  • RFP Document (required): The Request for Proposal PDF
  • Concept Document or Text (required): Your initial project concept
  • Reference Proposals (optional): Previously successful proposals for guidance
  • Supporting Documents (optional): Additional context materials
All uploaded documents are automatically vectorized using Amazon Titan Embed v2 for AI-powered retrieval during generation.

​
RFP analysis

Once documents are uploaded, trigger RFP analysis: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/analyze-rfp \
  -H "Authorization: Bearer YOUR_TOKEN"
The system extracts:
  • Requirements: Technical, organizational, and eligibility requirements
  • Evaluation Criteria: Scoring framework and weighting
  • Deliverables: Expected outputs and milestones
  • Submission Guidelines: Format, page limits, and deadlines

​
Concept analysis

After RFP analysis completes, analyze your concept: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/analyze-concept \
  -H "Authorization: Bearer YOUR_TOKEN"
The AI identifies:
  • Sections needing elaboration to meet RFP requirements
  • Gaps between your concept and RFP expectations
  • Strengths to emphasize
  • Suggested improvements

​
Step 2: Concept review

​
Review generated analysis

The system presents sections identified for elaboration: 
{
  "sections_needing_elaboration": [
    {
      "title": "Climate Adaptation Strategies",
      "rationale": "RFP requires specific climate resilience approaches",
      "selected": true
    },
    {
      "title": "Monitoring and Evaluation Framework",
      "rationale": "RFP mandates detailed M&E methodology",
      "selected": true
    }
  ]
}

​
Select and comment

Toggle section selection and add comments: 
const updateEvaluation = {
  selected_sections: [
    { title: "Climate Adaptation Strategies", selected: true },
    { title: "Monitoring and Evaluation Framework", selected: false }
  ],
  user_comments: {
    "Climate Adaptation Strategies": "Focus on drought-resistant crop varieties"
  }
}

​
Generate concept document

Generate an enhanced concept document incorporating your selections: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/generate-concept-document \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"concept_evaluation": ...}'
The result is a markdown document with:
  • Original concept sections
  • AI-generated elaborations for selected topics
  • Integration with RFP requirements

​
Step 3: Structure and workplan

​
Generate proposal structure

Create a structured outline based on RFP requirements: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/generate-structure-workplan \
  -H "Authorization: Bearer YOUR_TOKEN"
The AI generates:
  • Section hierarchy: Main sections and subsections matching RFP format
  • Page allocation: Estimated pages per section based on RFP limits
  • Workplan: Timeline, activities, milestones, and deliverables
  • Budget framework: Budget categories aligned with RFP requirements

​
Review and customize

The structure is returned as a JSON object: 
{
  "sections": [
    {
      "title": "Executive Summary",
      "subsections": [],
      "estimated_pages": 1,
      "guidance": "Concise overview highlighting key points"
    },
    {
      "title": "Technical Approach",
      "subsections": [
        "Methodology",
        "Work Plan",
        "Risk Management"
      ],
      "estimated_pages": 8
    }
  ],
  "workplan": {
    "phases": [...],
    "milestones": [...]
  }
}

​
Step 4: Proposal review

​
Generate draft template

Create a Word document template with your structure: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/generate-proposal-template \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -o proposal-template.docx
The template includes:
  • All sections and subsections from your structure
  • Placeholder guidance text for each section
  • Properly formatted headers and styles

​
Analyze draft feedback

Once you’ve written content, upload and analyze your draft: 
curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/documents/upload-draft \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "[email protected]"

curl -X POST https://api.igad-hub.com/api/proposals/{proposal_id}/analyze-draft-feedback \
  -H "Authorization: Bearer YOUR_TOKEN"
Receive detailed feedback: 
{
  "overall_score": 7.5,
  "alignment_with_rfp": 8.0,
  "technical_soundness": 7.0,
  "clarity": 8.0,
  "strengths": [
    "Strong methodology section",
    "Clear timeline and milestones"
  ],
  "improvements": [
    "Expand risk mitigation strategies",
    "Add more detail on monitoring indicators"
  ],
  "section_feedback": [
    {
      "section": "Technical Approach",
      "score": 8.5,
      "comments": "Well-structured and comprehensive"
    }
  ]
}

​
Data persistence

All workflow data is stored in DynamoDB:
  • Proposal metadata: Status, timestamps, user information
  • Analysis results: RFP analysis, concept analysis, structure
  • Generated documents: Concept document v2, templates
  • User selections: Section choices, comments, preferences

​
Next steps

Document management

Learn about document upload, storage, and vectorization

API reference

Explore the complete Proposal API

Admin prompts

Customize AI prompts for each workflow step

Architecture

Understand the underlying system architecture

Build docs developers (and LLMs) love

Get started for freeTalk to us