Leaf Technology own product
What it is
A public enterprise AI assistant demo: sync documents from a NAS folder into a citable knowledge base, look up operations data, and hold notifications for human confirmation—try without signing up.
Who it’s for
- ・Companies whose policies live on a NAS and want natural-language lookup
- ・Teams that need AI to query projects / receivables with permission control
- ・Leads who want Agents to require human confirmation for critical actions
Background
Company files often sit in NAS folders for admin, HR, or care. A general chat model cannot read them, and it also cannot safely query a database or send notices. Leaf Technology’s public Workspace puts NAS sync, cited answers, and pending actions in one assistant.
Problem
A single chat API cannot scan a NAS, cite a page, query operations data, or avoid claiming that an email was sent without a person confirming.
Solution
The knowledge base scans a designated NAS folder and syncs changes incrementally. Policy answers cite filename and page. Projects and overdue receivables use whitelist tools. Collection notices stay drafts until a person confirms; the public demo never sends email.
Architecture
- ・NAS folder scan and incremental sync (added, updated, unchanged, deleted)
- ・Enterprise knowledge-base index with citations (filename, page)
- ・Whitelist tools: search, query projects, query overdue receivables, create drafts
- ・Human confirmation gates and audit logs
- ・No-signup public demo UI
Core features
- ・Sync admin, HR, and care documents from a NAS folder
- ・Incrementally index new or edited files without rebuilding the whole library
- ・Company policy Q&A with knowledge-base filename and page
- ・Query in-progress projects and overdue receivables (demo data)
- ・Collection-notice drafts and post-confirmation audit
- ・Public demo uses a local folder as the NAS; production can mount a real NAS without changing retrieval
- ・Model cannot compose SQL—only call controlled tools
- ・Public demo https://leafflow-ai.vercel.app/knowledge
Technical challenges
Design “can answer,” “can query,” and “can act” separately so the model doesn’t merge query and execution into an untestable black box.
How we solved it
Documents via RAG, data via tools, actions via human confirmation; demo boundaries spell out fictional data and no real outbound email.
Results
A publicly demonstrable enterprise AI assistant architecture matching real delivery capability in RAG, API / database integration, and controlled Agents.
FAQ
How does Leaf AI Workspace use NAS files?
The public demo scans a designated folder, indexes admin, HR, and care documents, and incrementally syncs new or edited files. Answers cite filename and page. Production can point that path at a mounted NAS without changing retrieval.
How is Leaf AI Workspace different from general ChatGPT?
A general chat model does not know the company files on a NAS. Workspace searches the indexed documents with sources, reads demo operations data via whitelist tools, and waits for human confirmation before notification-class actions.
Will AI send collection emails directly?
Not in the public demo. The system creates a draft and pending-confirmation flow; after confirm it only writes an audit log—no email is sent.
Related services
If you have a similar Web product, quiz platform, or AI productization need, start with Custom Web Systems,AI System Integration, or request a development assessment.