--- title: "Automatic Document Archiving in 14 Days | JustAutomate" description: "An invoice, a contract or a scan files itself in the client folder on Drive or SharePoint: one naming pattern, access by role. First stream in 14 days." url: "https://justautomate.ai/en/automations/automatic-document-archiving/" lang: en --- Finance and documents # Automatic document archiving no more asking where that invoice is An invoice from your invoicing system, a signed contract from Autenti or DocuSign, a scan from an inbox, a report from a spreadsheet. The workflow catches the document the moment it is created and files it in Google Drive or SharePoint under the right client, year and type. One naming pattern, access by department, a notification to the person who needs to see it. Only where the file lands changes. Book a free consultation [See how the workshop works](https://justautomate.ai/en/services/process-workshop/) - 19%of a knowledge worker's week goes on searching for and gathering information, close to one day a week. McKinsey Global Institute, 2012 - 14 daysfrom the workshop to the first stream of documents, usually invoices, filing itself. The whole rollout takes 2 to 6 weeks - 5 to 10 secondshow long BOND, our company AI agent, takes to answer a question about a document out of the archive and the CRM The problem ## Why does finding a contract take longer than signing it? The contract is in a rep's inbox, in a thread with fourteen replies. The purchase invoice came to the general office address and somebody saved it to their desktop. The handover note sits in Teams, in a project channel that was closed months ago. The scan with the client's signature is in the Dropbox of a person who left in March. When the accountant asks for an attachment to an invoice from last year, three people start searching and a fourth emails the client for a duplicate. This is not carelessness. Everyone saves where they happen to be working, and every system produces documents its own way: the invoicing tool issues a PDF, Autenti emails back a signed contract, Google Sheets generates a report, the scanner sends to a mailbox. The rule of client, year, document type does exist, but it lives in the office manager's head, and it breaks every time somebody has five minutes before a meeting. The cost is countable. According to the [McKinsey Global Institute report The social economy from 2012](https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-social-economy), a knowledge worker spends 19% of the working week searching for and gathering information, close to one day a week. Then there is the law. How long you keep an accounting record, a contract or an HR file follows from national legislation and from the contracts you signed, and it differs by document type. [GDPR 2016/679, Article 5(1)(e)](https://eur-lex.europa.eu/eli/reg/2016/679/oj) sets the other half of the rule: personal data is kept no longer than it is needed for the purpose. A former employee's mailbox satisfies neither condition. - The accountant waits for an invoice attachment while three people search their inboxes - A signed contract exists only in the email of a rep who has left the company - The same file sits in four versions and nobody knows which one is final - An audit asks for documents from three years back and collecting them takes days, not minutes Why now ## What changed, so this runs without the team changing habits? Every system that produces documents now has a webhook or an API. Fakturownia and iFirma report a new invoice a second after it is issued, Autenti and DocuSign send back a signed contract together with its status, Gmail and Outlook filter attachments by sender and subject, Google Sheets exports a report on a schedule, and purchase invoices can be pulled from the structured e-invoicing channel your country runs. Subiekt and Comarch Optima are local programs, so they are reached through a connector sitting inside your own network. The document can be caught the moment it exists, before anyone saves it somewhere. On the other side, Google Drive and SharePoint expose APIs for folders, names and permissions. A workflow in [n8n](https://docs.n8n.io/) or Make creates the folder Client / 2026 / Invoices, names the file to the agreed pattern and grants access to a group in Google Workspace or Microsoft 365\. The rule leaves the office manager's head and becomes a rule in a system. The team works in the same tools as before and nobody learns a new program. Process first. Tools second. The third change is the AI agent. A workflow will put the file in the right place, but it will not answer whether you have a signed contract with company X and when it expires, or which August purchase invoices are missing. BOND, our company AI agent, answers that out of the archive and the CRM in 5 to 10 seconds, with no folder hunting and no asking three people on Slack. What changes ## What does the team get once documents file themselves? - ### Every document in one place Invoices, contracts, handover notes, quotes and scans land in Google Drive or SharePoint the moment they are created, in a folder by client, year and type. Nobody saves them by hand and nobody decides where they go. One folder structure for 5 kinds of source: invoicing system, electronic signature, CRM, spreadsheets, inbox - ### A file name you can actually search A pattern such as INVOICE\_2026-03\_ClientXYZ.pdf, agreed in the workshop and applied to every file without exception. The search box in Drive or SharePoint finds the document by type, client and month. One naming pattern instead of scan0001.pdf and contract\_final\_v3\_revised.docx - ### Access by role, not by who sent it Finance and the board see invoices, sales and the lawyer see contracts, HR files stay with HR. Permissions are granted by the workflow from a group in Google Workspace or Microsoft 365, not by a person clicking share. Access removed with the account, with no file by file review of shares after somebody leaves - ### A notification to the person who needs to see it A new purchase invoice: a Slack message to the accountant. A signed contract: an email to the account owner and a link in the CRM. A handover note: Teams to the project manager. Everyone else gets nothing. The message arrives seconds after the file does, with a link straight to its place in the archive - ### Ready for an audit, and for the retention rules you have to keep Accounting records sit in a structure you can hand over in a day rather than in weeks. Date, counterparty and document type are in the name and in the metadata, and the retention period is set upfront for each type, from the rule that actually applies to it. Retention set per document type in the workshop, from national law and from your contracts, instead of living in one person's memory - ### Questions about documents asked in a plain sentence Do we have a signed contract with company X and until when. Which August invoices are missing from the archive. Show me the handover notes from project Y. The AI agent answers from the archive and the CRM, in English or in Polish. An answer in 5 to 10 seconds, with no searching through folders and inboxes And in your company ## What if our documents are everywhere? Four sentences we hear most often, and what we answer to each of them. If you are after a neighbouring piece, there are separate pages for [automatic contract generation](https://justautomate.ai/en/automations/automatic-contract-generation/), for [automatic invoicing from a CRM](https://justautomate.ai/en/automations/automatic-invoicing-from-crm/) and for [document automation](https://justautomate.ai/en/services/document-automation/) as a whole. - We are 8 people on Google Drive, why would we automate this? If documents come out of one system and there are 20 a month, folder discipline is enough, and we say so on the consultation. Automation starts paying off when documents arrive from at least 3 systems, or the first time somebody leaves and you go looking for a contract in their mailbox. We come back to it then. - Our documents are in inboxes, in Dropbox and on desktops, there is no way to sort that out. We do not sort out the past on day one. The workflow starts with documents created from today, and that is the first result after 14 days. The back archive is a separate step: a migration out of Dropbox and the inboxes in batches, by client and year, with duplicates caught on a checksum. In the workshop you decide how far back we go, the full retention period for a document type or only the current year. - We run Microsoft 365 and SharePoint, not Google. The mechanism is the same: document libraries instead of Drive folders, metadata in columns, permissions from Microsoft 365 groups, notifications on Teams instead of Slack. A workflow in n8n or Make talks to SharePoint over the API exactly as it does to Drive. The place changes, the rules do not. - We are worried the workflow will file an invoice under the wrong client. The workflow identifies the client by the VAT number read from the invoice, by the client number from the CRM or by the sender address. When confidence is low the file goes to an unassigned folder with a notification, not to a random client. For the first 2 weeks it runs alongside the old way, so you compare results before switching anything off. Invoice sources Fakturownia, iFirma, sevDesk, Comarch Optima and Subiekt through a connector in your network, national e-invoicing channels Electronic signature Autenti, DocuSign CRM HubSpot, Pipedrive, Salesforce, LiveSpace Inboxes, scans and spreadsheets Gmail and Outlook attachments, a network scanner sending to an email address, Google Sheets, Dropbox as a migration source Archive and notifications Google Drive, SharePoint, Slack, Microsoft Teams, email Automation layer n8n, Make, custom integrations over the API ## Document types, where we catch them and where they land The documents that most often make up a B2B company archive, the source the workflow catches them from, and the rule it files them by. The rollout order is settled in the workshop. | Document type | Where the workflow takes it from | Filing rule | | -------------------- | -------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------- | | Sales invoice | Fakturownia and iFirma over a webhook or the API, Subiekt and Comarch Optima through a connector inside your network | Client / year / Invoices; named INVOICE\_YYYY-MM\_Client\_number.pdf | | Purchase invoice | an attachment from Gmail or Outlook, a pull from the e-invoicing channel | Supplier / year / Costs; the supplier recognised by the VAT number in the document | | Contract and annex | Autenti or DocuSign once signing completes | Client / Contracts; the link and the expiry date go back into the CRM | | Quote | the PDF from the CRM or from a quote generator | Client / year / Quotes; every version numbered, the old one is never overwritten | | Handover note, order | a form, an email from the client, Teams or Slack | Client / Project / Handovers; a notification to the project manager | | Recurring report | Google Sheets, BigQuery, an export from a tool | Client / year / Reports; the period in the name, REPORT\_2026-08\_Client.pdf | | HR document | the HR system, a scan, an onboarding form | Employee / type; access limited to HR and the line manager | The structure and the naming pattern come out of the workshop. A first rollout usually covers 2 to 3 document types, most often invoices and contracts, and we add the rest after the first 30 days. How we do it ## Five steps from documents everywhere to an archive that keeps itself Each step ends with something you can judge before you decide on the next one. 1. ### Workshop We write down which documents get created, in which systems, who goes looking for them later and where they land today. We check who has access to what and what the retention rule is for each type, taking it from the legislation that applies to you and from your contracts. The output: a source map, a folder structure, a naming pattern and an access matrix. 2. ### Rule design For each document type: where we take it from, how we identify the client, where we file it, who gets access and who gets the notification. We settle what happens when the client cannot be identified. The tool choice: n8n, Make or an integration straight over the API. 3. ### First stream in 14 days We start with one type, usually invoices, because they have the simplest rules and the most people hunting for them. Within 14 days of the workshop invoices file themselves, alongside the current way of working, so nothing is lost if a rule needs correcting. 4. ### More types and the back migration We add contracts, handover notes, quotes and reports. If that is what you decided, we move the back archive out of Dropbox and the inboxes in parallel, in batches by client and year, with duplicates caught on a checksum. 5. ### Switch and observation We turn off manual saving and turn on error monitoring. For 30 days we watch the unassigned folder: how many documents landed there and why. The office manager learns to change the rules alone: a new client, a new document type, a new person in a group. Before and after ## How the day changes for the office manager and the accountant ### Before - Answers three where is questions before lunch and digs through the inbox for attachments - Asks the client for a duplicate invoice, because the original sat in the email of somebody on holiday - Collects purchase invoices off five desktops before the month closes - Reviews a leaver's shares file by file after they hand back the laptop ### After - The invoice is in the client folder before anyone has opened it - A signed contract has a link in the CRM and an expiry date in the calendar a minute after signing - Month close starts from a complete folder, not from an email asking people to send their invoices - Somebody leaving means closing an account, and the documents stay in the company structure The first invoices file themselves 14 days after the workshop. We judge the folder structure and the naming pattern after 30 days, on what actually arrived, including what is sitting in the unassigned folder. For the person signing this off ## What the finance director or the owner gets out of it The question of where the contract with company X is stops reaching you, because the answer is in one place and everyone entitled to it can see it. When an audit comes, you hand over the documents in a day instead of spending a week collecting them from the mailboxes of people who no longer work here. The second thing is a risk that appears in no spreadsheet today. A contract in a private mailbox, an invoice on a desktop and a scan in the Dropbox of somebody who left are documents the company formally does not hold. After the rollout you have a structure, group permissions and a history of who added what. That is an argument in front of an auditor and in front of a board. - The full set of accounting records in one structure, with the retention period set per document type - Access by role, withdrawn together with the account, with no manual clean up of shares - The first stream runs after 14 days and the old way stays until the tests are finished - A new client, document type or person: you change the rule yourself, without raising a ticket with us What it costs ## The workshop has a published price, the rollout is quoted after it The process workshop costs **EUR 3,500 net per week**. For archiving one week is usually enough, because the process is predictable: a document appears, it has to be identified, named, filed and shared. We quote the rollout after the workshop, once it is clear how many sources and document types are involved and whether we are moving the back archive. You pay the cost of running the workflow directly to the tool vendor, meaning n8n or Make, or you pay nothing at all when n8n runs on your own infrastructure. Storage is your existing Google Workspace or Microsoft 365 subscription. Where the AI agent is part of the scope, model queries cost the vendor EUR 0.005 to 0.07 each. No commission from us on any of it. - ### Process workshop EUR 3,500 net per week, on archiving usually one week A self-contained piece of work. It ends with a source map, a folder structure, a naming pattern and an access matrix you can roll out without us. - A map of the flow you have today: which documents, from which systems, where they land and who looks for them - Folder structure and naming pattern written down together with the exceptions - An access matrix by role and the retention rule for each document type - Tool recommendation: rules in Drive or SharePoint, n8n, Make or the API Start with the workshop - Most popular ### Workshop and archiving rollout Quoted after the workshop scope follows from the number of sources and document types The option most companies choose. The first stream of documents after 14 days, the whole rollout in 2 to 6 weeks from the workshop, tested on your own files. - Everything in the workshop - A workflow in n8n, Make or over the API, wired into the document sources and into Drive or SharePoint - Client identification, naming, group permissions, notifications and the unassigned folder - A parallel rollout and a switch with error monitoring - Training for the office manager on changing the rules alone Talk about a rollout - ### Archiving and an AI agent for documents Quoted after the workshop for companies where the where is it question comes up a dozen times a day Everything in the rollout plus BOND, which answers questions about documents in a plain sentence, out of the archive and the CRM. - Everything in the archiving rollout - An AI agent wired into the archive and the CRM: do we have the contract, until when, which invoices are missing - Alerts on contracts expiring within 30 days and on invoices with no client assigned - Access for finance, sales and the board, with the same permissions they have in the archive Ask about the agent If the workshop shows that documents come out of one system and all you need is a tidy folder and a mail filter, we say so and show you how. The workshop ends with a document: a folder structure, a naming pattern and an access matrix that stand on their own without us. ## Frequently asked questions What is automatic document archiving? It is a workflow that catches a document the moment it is created, an invoice from the invoicing system, a contract signed in Autenti or DocuSign, an email attachment, a report from a spreadsheet, and saves it in Google Drive or SharePoint in a folder by client, year and type. It names the file to one pattern, sets access by role and notifies the person who needs to see it. A human only steps in when the workflow cannot identify the client. How long does automatic document archiving take to implement? One week of workshop, then the first stream of documents, usually invoices, after 14 days. The whole rollout, with contracts, handover notes and reports, takes 2 to 6 weeks depending on the number of sources. Migrating the back archive is a separate stage and depends on how far back you want to go. Do we have to move off Google Drive or SharePoint to something else? No. The archive stays where the team already works: Google Drive in Google Workspace or SharePoint in Microsoft 365\. What changes is who saves the files and by which rule, not the place. Nobody learns a new program, and the search box you use today starts finding things, because the files finally have names that mean something. How does the workflow know which client a document belongs to? By the VAT number read out of the invoice, by the client or deal number from the CRM, by the sender email address and by the company name matched against your client list. When none of those gives certainty, the file goes to an unassigned folder with a notification to the office manager. A document never lands under a random client and it never disappears. What about the documents we already have in inboxes, Dropbox and on desktops? From day one the workflow handles documents created as you go. The back archive is moved separately, in batches by client and year, with duplicates caught on a checksum so the same scan pulled from three mailboxes does not land in the folder three times. In the workshop you decide how far back we go, the full retention period for a document type or only the current year. How long do we have to keep invoices, and does a cloud archive satisfy that? The retention period is not something we set. It follows from the legislation that applies to your company and from the contracts you signed, and it differs by document type: accounting records, contracts, HR files and records with personal data each have their own clock. Our job is to make the system hold those periods instead of one person remembering them. Article 5(1)(e) of GDPR 2016/679, the storage limitation principle, also sets an upper bound on personal data. In the workshop we design the folder structure, the naming and the access rules so that each type carries its own retention period and a document can be produced on request. Is the data in our documents safe when it passes through n8n or Make? The workflow moves a file between two systems you already run, and it gets access only to specific folders, not to the whole drive. n8n can run on your own infrastructure, in which case the document never leaves your environment. HR files and anything else carrying personal data get their own rule: a narrower access list and a retention period agreed in the workshop, in line with GDPR. What does the AI agent add on top of archiving? Archiving puts the file in the right place. The AI agent answers the questions a folder does not understand: do we have a signed contract with company X and until when, which August purchase invoices are missing, show me the handover notes from project Y. BOND answers out of the archive and the CRM in 5 to 10 seconds, with the same permissions that person has in the archive itself. Rolling out the agent takes around 2 months and it pays off when the where is it question comes up a dozen times a day. ## Let us find out where your documents disappear today The free consultation takes 30 minutes. You tell us which systems the documents come from and who goes looking for them afterwards, even if the rule for where to save things exists only in one person's head. We show you which streams can be automated straight away and which need the structure sorted out first. You do not have to give access to any system or drive at this stage. We run at most two rollouts at a time, so that someone is actually sitting on each of them. [Book a free consultation](https://justautomate.ai/en/contact/) **PS.** If somebody at your company wrote has anyone got that invoice on Slack this week, that is exactly this problem. A document sitting in someone's mailbox does not formally belong to the company, and it cannot be found when that person is on holiday or has left. After the rollout every file has its place before anyone gets to ask about it. ## Related - [Document automation](https://justautomate.ai/en/services/document-automation/) - [Automatic contract generation](https://justautomate.ai/en/automations/automatic-contract-generation/) - [Automatic invoicing from CRM](https://justautomate.ai/en/automations/automatic-invoicing-from-crm/) - [n8n automation](https://justautomate.ai/en/services/n8n-automation/) - [BOND, business AI agent](https://justautomate.ai/en/services/bond-business-ai-agent/) - [Process workshop](https://justautomate.ai/en/services/process-workshop/) --- **JustAutomate** · contact: contact@justautomate.ai · form: https://justautomate.ai/en/contact/