Own MinerU.

What you'd own.

Every artifact below transfers to your accounts on day one. The whole engine.

• ●MinerU brand, domain, and social proof with 30k GitHub stars.
• ●Landing page, product assets, marketing copy and design system.
• ●Open-source MinerU repository and active community stewardship.
• ●API backend codebase, parsing models, and document processing pipeline.
• ●Customer list, email sequences, and GTM playbooks for developer communities.
• ●Founding team conversations and competitive analysis versus LlamaParse and Unstructured.
• ●Year 1 financial model and bottleneck analysis for strategic decisions.

What Roll Digital builds for you.

Custom scope to take MinerU from MVP shell to operating business.

• ▸Production FastAPI backend with async job processing and queue management.
• ▸Stripe integration for 3-tier pricing and subscription management.
• ▸Multi-tenant auth with email, Google sign-in, and API key management.
• ▸Postgres database with document versioning and processing history.
• ▸Webhook system for batch processing callbacks and external integrations.
• ▸Custom domain provisioning, SSL termination, and enterprise deployment.
• ▸30-day customer success and onboarding program for early paying users.
• ▸Monitoring, logging, and alerting for production parsing reliability.

Why this could work.

The core value is real: RAG builders spend weeks converting PDFs to markdown manually. A fast API solves a genuine pain. The GTM is validated: free-to-paid funnels work for dev tools. TAM is substantial: 20k companies actively building RAG pipelines. The challenge is incumbent competitors and self-hosted alternatives already capturing the market. Winning requires strong product quality and community trust.

What you'd risk.

The biggest risk is that the open-source MinerU repo with 30k stars will always undercut you. Most target customers self-host for free rather than pay. LlamaParse and Unstructured.io are well-funded incumbents with enterprise sales teams. API providers like OpenAI, Anthropic, and Google are shipping native file parsing into their APIs within 12-18 months, commoditizing the category. Document quality variance silently kills retention before users report issues. This is a high-risk, long-runway bet.

Who this fits.

Ideal owner is an experienced SaaS operator or technical founder with an existing audience in AI/ML/RAG communities: Twitter, GitHub, Substack, mailing lists. Must have at least 50k runway and deep familiarity with document processing or LLM infrastructure. Requires comfort with 13% success odds and patience for Year 2-3 before break-even. This suits someone pivoting from an agency or spinning out from a larger AI platform.

Timeline.

From contract signing to operating business.

1. Day 1
   Brand transition complete, financial models reviewed, founder debrief.
2. Week 1
   FastAPI backend deployed, Postgres schema designed, async processing queued.
3. Week 3
   Free-tier API live, 10-doc-per-day limit, initial beta signups.
4. Week 6
   Stripe billing integrated, public launch, SEO optimized, HN post.
5. Day 90
   50-150 paying customers acquired, churn analysis begun, Year 2 roadmap.

Pricing.

Three ways in, depending on how much you want to build yourself.