Knowledge Base
Your knowledge base is the foundation of great proposals. Upload past proposals, case studies, certifications, and more — the AI uses these documents to write accurate, evidence-backed content.
How It Works
Our knowledge base uses advanced retrieval-augmented generation (RAG) to ensure proposals are grounded in your real capabilities and past work.
Upload
Upload documents in PDF, DOCX, or TXT format
Chunk & Embed
Documents are split into semantic chunks and converted to 768-dimension vectors
Hybrid Search
Semantic vector search + keyword matching find the most relevant chunks
LLM Re-rank
AI re-ranks results to select the best evidence for each requirement
Document Types
Choose the right document type to help the AI understand and use your content effectively.
Case Studies
Past client success stories with metrics and outcomes
Past Proposals
Previous RFP responses and winning proposals
Company Overviews
Company background, mission, and capabilities
Team Bios & Resumes
Staff qualifications, experience, and credentials
Service Descriptions
Detailed descriptions of your service offerings
Certifications & Compliance
Industry certifications, compliance docs, security policies
Rate Cards
Hourly rates, T&M pricing, and cost structures
SLA Templates
Service level agreements and performance guarantees
Technical Architecture
Infrastructure designs, system architecture docs
Network Diagrams
Network topology, data flow, and integration diagrams
Product Documentation
Feature specs, user guides, release notes
API Documentation
REST/GraphQL API references, webhooks, SDKs
Integration Guides
Third-party integrations, API setup, data migration
Upload Best Practices
Use specific document types rather than "Company Overview" for everything. This helps the AI understand context and retrieve the right content.
Include concrete metrics and outcomes in case studies. Numbers like "reduced downtime by 40%" or "saved $200K annually" make proposals more compelling.
Keep documents focused — split large docs into topic-specific files. A 50-page services catalog is better as 5 separate service descriptions.
Update documents regularly to reflect current capabilities. Outdated docs can lead to inaccurate proposals.
Name files descriptively for easier organization. Use names like "Case_Study_Healthcare_HIPAA_2025.pdf" instead of "document1.pdf".
How AI Uses Your Documents
During proposal generation, the AI performs hybrid search across your knowledge base for each requirement in the RFP. It combines semantic vector search (understanding meaning) with keyword matching (exact terminology), then uses an LLM to re-rank results and select the most relevant evidence.
This RAG (Retrieval-Augmented Generation) approach ensures proposals are grounded in real facts from your documents, not generic AI hallucinations. The more comprehensive your knowledge base, the more accurate and compelling your proposals will be.
Want to understand the full proposal generation workflow? Read the RFP Workflow guide.
Ready to get started?
Create your free account and generate your first AI-powered proposal in minutes.
Start Free