Turn proprietary deal materials into prospectus-ready drafts.
AI Prospectus is a modular drafting system for sponsor-counsel workflows: evidence preparation, section generation, verification notes, and Word export in one controlled workspace.
Disclosure drafting pipeline
Evidence first, generation second
Ingest proprietary materials
Upload issuer spreadsheets, JSON records, diligence outputs, and internal workpapers into a controlled drafting workspace.
Prepare evidence
Agent1 separates narrative text from structured facts, maps content to prospectus sections, and preserves source pointers.
Generate regulated drafts
Agent2 drafts section-by-section with verification notes, missing-data flags, and Word export for professional review.
Built for proprietary document generation
The project started with IPO prospectus drafting, but the architecture now separates the AI module, frontend workspace, platform tools, knowledge module, and pipeline module so domain-specific drafting workflows can evolve independently.
Private data workflow
Runtime files can stay in a local workspace while the codebase remains clean and modular.
Section-aware drafting
Requirements, evidence retrieval, verification, and output bundles are organized around disclosure sections.
Desktop-ready packaging
Downloadable builds support internal review and distribution without exposing issuer files to a public SaaS.
Extensible modules
AI, frontend, platform, evidence layer, resources, and extraction pipelines now have clear homes.
Inside the drafting engine
A private evidence graph organizes the facts before the model writes.
Users work with upload, draft, revise, and export. Behind that surface, AI Prospectus builds a structured evidence layer that keeps source materials, extracted facts, section requirements, and missing-input signals aligned throughout generation.
Source anchoring
Facts, narrative passages, and extracted tables stay connected to the files they came from.
Section routing
Evidence is classified against the prospectus section map before drafting begins.
Gap detection
Missing inputs and weak coverage are surfaced as review signals instead of hidden in the draft.
Internal evidence layer
Source material to verified disclosure structure
Issuer files
PDF, DOCX, XLSX, JSON
Extracted facts
values, dates, entities
Section map
Business, Risk, MD&A
Source pointers
file, table, page, row
Coverage gaps
missing inputs flagged
Draft plan
evidence-backed outline
From prospectus drafting to private document generation
AI Prospectus is designed to grow beyond a single drafting template. The same modular structure can support issuer questionnaires, diligence summaries, regulated filings, board papers, and other proprietary documents where traceability and review discipline matter.
Demo and partnership enquiries
For pilots, sponsor-counsel workflows, or private deployment discussions, request a demo and include your target document type.
Request demoSource and releases
The public site links to GitHub releases for downloadable builds while the protected workspace remains separate from the marketing surface.
View GitHubDownload the app or request access to the workspace.
Use the website for product discovery, then move to a protected workspace for controlled drafting.