The Document Concierge — automated filing, parsing & cited answer search

What you get

Stop losing documents and re-answering the same questions. Let AI file every document and answer from them. Incoming documents — contracts, invoices, reports, scanned PDFs — pile up in inboxes and shared drives until no one can find anything. This agent watches the folder or inbox you point it at, reads and classifies each new document, and files it where it belongs (renamed and tagged by rules we agree on). Then your team chats with an assistant that parses across those documents and answers questions with citations back to the exact source, so every answer is verifiable. AI retrieval is good but not infallible, which is exactly why every answer cites its source. WHO IT'S FOR: Teams drowning in inbound documents — operations, finance, or back-office (10+ employees) — who need filing handled automatically and answers they can trust. Not a fit if your documents already live in a well-tagged system you're happy with. WHAT'S INCLUDED: - A watched intake (a Google Drive or SharePoint folder, or an email inbox) that auto-parses and classifies every new document - Auto-filing into your folder structure, renamed and tagged by rules we agree on - A private chat assistant that answers questions across your documents with source citations - A private vector store used only for your assistant - Update scheduling so newly filed documents become searchable automatically NOT INCLUDED: Building your folder taxonomy from scratch (we organize into the structure you already use) or bulk data-entry of paper records not provided digitally. CAPACITY: The base build covers one intake source and one document set scoped at kickoff. Additional sources or very high volumes are quoted before anything is billed. HOW YOUR DATA FLOWS: Your documents live in a private vector store used only for your assistant. Queries are processed by the LLM provider under its API data-use terms; current major-provider API policies do not use API inputs for training. We walk through exactly where your data flows at kickoff so there are no surprises. THIRD-PARTY COSTS YOU PAY DIRECTLY: LLM and document-parsing API usage runs typically $30 to $120 a month for a 10-person team at normal volume, and hosting runs roughly $10 to $25 a month. Both are billed to your own accounts. DELIVERY: Live within 10 days of you granting access to your intake folder or inbox and document set. Includes 30 days of post-handoff support and one revision round. THE RETAINER ($600/mo) After handoff, $600 a month covers monitoring the filing and chat agent, fixing parsing or retrieval issues, and tuning classification rules as your document mix changes. It bills monthly starting at handoff. Cancel anytime: tell us and you aren't invoiced again. HOW IT WORKS: 1. Book the implementation call. We review your intake source and document set and agree on what's in scope, including volume, so ingestion is bounded before anything is billed. 2. Fixed scope and price in writing; private checkout link after you approve it. 3. After the call you grant access to your intake folder or inbox (we send a short checklist the same day), then we wire the auto-filing, ingest, test answer quality and filing accuracy against real documents, and hand off. We confirm your stack on the call. If we can't deliver what we scoped, we say so and refund in full.