Hackathon project built for OxBuild by Oxlo.ai
https://oxlovision.vercel.app/
Oxlo Vision was built to demonstrate a complete, real-world AI workflow in one product:
- Input: upload digital or scanned PDFs
- Processing: extract text + OCR fallback
- Intelligence: orchestrate Oxlo.ai models through a unified backend
- Output: summaries, key points, mind maps, concept maps, markdown, and diagrams
For judging, the fastest validation path is to test the live product directly:
Oxlo Vision transforms PDF documents into actionable knowledge for students, researchers, and developers:
- Extracts text from standard PDFs and scanned PDFs (OCR fallback)
- Generates concise summaries and key points
- Builds mind maps and concept maps from real extracted content
- Produces markdown and AI-ready skill outputs for developer workflows
- Supports chat-based exploration over uploaded document content
Most PDF workflows are slow and fragmented:
- Scanned documents cannot be searched or copied easily
- Users spend too much time reading large files manually
- Insights are trapped in unstructured text instead of reusable outputs
- Teams need multiple tools to summarize, map ideas, and document findings
Oxlo Vision solves this by converting raw PDFs into structured, reusable intelligence in a single flow.
Oxlo Vision is built with these product objectives:
- Reduce time-to-insight from long documents
- Make scanned PDFs usable through OCR and text extraction
- Deliver high-quality summaries and key findings for fast decision-making
- Generate visual understanding artifacts (mind maps, concept maps, diagrams)
- Provide developer-ready outputs (Markdown and skill files for AI assistants)
- Offer a deployable, real-world solution judges can test immediately
This project is designed to demonstrate practical value, technical execution, and product readiness:
- Fast end-to-end experience: upload PDF -> extract -> analyze -> visualize
- Real AI orchestration through Oxlo.ai model ecosystem
- Clear developer-focused outputs (markdown, diagrams, skills)
- Production-style deployment and API proxy strategy
flowchart LR
U[User]
FE[Oxlo Vision Frontend\nReact + TypeScript]
API[Proxy Layer\n/api + /v1]
BE[Backend\nMicronaut + JDK 21]
OX[Oxlo.ai Platform\nLLMs + Vision + OCR + Embeddings]
U --> FE --> API --> BE --> OX
OX --> BE --> FE --> U
- PDF upload with drag-and-drop and file selector
- Text extraction with OCR support for scanned pages
- Document summary and key-point generation
- Mind map generation from extracted document semantics
- Concept map generation for structured understanding
- AI chat over uploaded document context
- Diagram generation support for technical documentation
- Product value: solves a real information extraction bottleneck
- AI integration depth: multiple Oxlo.ai capabilities in one flow
- Developer usefulness: exports and structures ready for AI-assisted coding workflows
- UX completeness: from raw PDF to insights in one interface
- Deployment readiness: publicly available and testable now
If you are evaluating Oxlo Vision, please start with the live demo above.
