Photograph a stack of paper contracts — get instant, cited answers, no scanner required.
A filing cabinet full of paper contracts is invisible to search — until now. Snap a photo, and it reads like it was always digital. Upload contracts, agreements, or reports and get an automatic summary, ask follow-up questions in plain conversation, search your entire document library at once, or compare two versions to see exactly what changed. Every answer tells you which page it came from — and every document stays entirely on your own computer.
Ubwiru Document Agent is in open beta (Beta v0.1). The free 14-day trial doubles as beta testing — your feedback shapes what ships in the 1.0 release.
Built-in OCR means a phone photo of a paper document becomes fully searchable, cited text — no scanner, no separate step.
Genuinely removes sensitive text from a copy of the file — it can't be recovered by copy-pasting or lifting a black box, unlike most "redaction" tools.
Every answer tells you exactly which page it came from — across one document or your entire library at once.
A section-by-section approach means nothing gets left out just because a document runs 100+ pages.
Get a plain-language explanation of what actually changed between two versions, not just a raw diff.
PDF, Word, PowerPoint, Excel, text, HTML, ebooks, images, and more — one library for everything, not just PDFs.
Every document you upload stays searchable going forward, building a library you can query as a whole.
Combine PDFs, split one into pages, or pull a specific range — handy file tools built right in.
Professions that spend real time reading dense documents where privacy matters most — lawyers, real estate agents, HR and recruiting, compliance and grant administrators, insurance professionals, healthcare administrators, and accountants.
Runs entirely on your own computer — here's what that computer needs.
| Requirement | Minimum | Recommended |
|---|---|---|
| Windows | 10 (64-bit) | 11 |
| macOS | 12 | 14+ |
| Linux | Ubuntu 20.04 LTS, Debian 11, or equivalent (glibc 2.31+) | Ubuntu 22.04 LTS or newer |
| Memory (RAM) | 8 GB | 16 GB+ for very long documents |
| Free Disk Space | 12 GB | 20 GB |
| Graphics Card (GPU) | Not required | Speeds up long-document summaries |
Python and the local AI engine are installed automatically during setup — nothing to install yourself. Internet is only needed during setup and updates; everything runs offline day to day.
Document Agent runs two AI models at once — a G-LOC G3 handles it well, and a G6 will feel noticeably snappier on longer documents.
PDF, Word, PowerPoint, Excel, scanned images, and more build your searchable library, entirely on your own computer.
Get a summary, ask a question with a cited page reference, or compare two versions side by side.
Pick your profession in Customization and Ubwiru knows what to watch for — unusual clauses, missing dates, coverage exclusions, and more.
Every summary and comparison is clearly marked to verify against the original — this tool is built to save you time finding the right place to look, not to replace reading what matters
Highlighting is best-effort — most passages are found and marked correctly, and the tool tells you plainly on the rare occasion one isn't
Every field's default template includes appropriate professional boundaries baked in from the start — for example, never offering legal or medical advice
No — everything runs entirely on your own computer, which is exactly why this tool exists for professions handling especially sensitive documents.
12 file types out of the box — PDF, Word, PowerPoint, Excel, text, Markdown, RTF, CSV, HTML, EPUB, OpenDocument, and images. Scanned or photographed pages are read automatically with built-in OCR.
Every summary and comparison is clearly marked to verify against the original — this is built to save you time finding the right place to look, not to replace reading what actually matters. Highlighting is best-effort, and the tool tells you plainly on the rare occasion a passage isn't found.
Yes — compare any two versions and get a plain-language explanation of what actually changed, not just a raw diff.
Yes — it genuinely removes the sensitive text from a copy of the file. It can't be recovered by copy-pasting or lifting a black box, unlike most "redaction" tools.
Yes — every Ubwiru AI tool runs entirely on your own computer using local, open-source AI models. Nothing is sent to our servers or any third party.
A one-year license, activated on one computer. It briefly needs internet during setup to activate, then runs fully offline. Moving to a new computer is a quick email to hello@ubwiruai.com.
Explore the rest of the Ubwiru AI lineup, or talk to us about what fits your business.
