Running a Digital Office

Back in 1999 I decided to go paperless. This was a significant challenge then as many bills were still being sent via post, and few offered digital alternatives. The challenge was doubly hard as I was (and am still) running Linux as my only operating system. These are a few of the things I had to do to make a digital office work and how things have evolved since then.

The Early Days

The choice of a Linux desktop meant that I had to find tools that were compatible with my operating system. This limited my options, but I was able to find open-source tools that met my needs. I used GIMP for image editing, LibreOffice for document editing, and PDFtk for PDF manipulation. My Canon printer had a built-in scanner, which I used to digitise paper documents. Canon printers and scanners are well supported by Linux. I used XSane for scanning documents. I also used rsync to maintain a continuous backup of my digital files.

Saving these scanned documents and bills as PDFs on my computer was a manual process that often led to large document sizes. (Disk space was more expensive back then.) OCR (Optical Character Recognition) was available, but not accurate, particularly as bills often used tables. I couldn't reliably search PDF images, so I had to rely on file names to find what I needed. At the same time, I started using GNUCash to manage my finances. All my transactions are recorded there, which helps me keep track of my expenses while also serving as an index to the scanned documents. Bills were managed monthly in a directory, the previous bill going into an archive directory. As my system was being backed up regularly using rsync, I had a backup of all my bills and financial records. Each financial year I would record that year's transactions and tax documentation using R Markdown with extracts from GNUCash. That financial year directory would also be archived onto a DVD and stored offsite. My accountant would send me a PDF of the tax return that required my input. For that purpose I would use LibreOffice to edit the PDF and add the required information. That often meant splitting the PDF into multiple pages, adding the information, and then merging it back together.

So I had a process that, while manual and time-consuming, worked. I had a digital office, albeit a very basic one.

Evolution of the Digital Office

More services transitioned to digital formats, and I adapted my processes accordingly. I started receiving bills via email, which simplified how I managed and stored them. I store these locally rather than leaving them on the email server. In addition, I could more easily annotate PDFs and add metadata (exiftool). While cloud storage was becoming more popular, I preferred to keep my documents local for privacy and control reasons.

While GNUCash offered invoicing capabilities, I found it more practical to use templates from LibreOffice.

As more documents are received digitally, I can easily search and organise them using metadata. This has made specific documents much quicker to find when needed. Fewer receipts and documents are now sent via post, but I still have to scan some documents that I receive in paper form. However, the process has become far less cumbersome with the advent of mobile scanning. Using my smartphone, I can quickly scan documents and save them as PDFs. OCR technology has also improved significantly, allowing me to extract text from scanned documents. This can then be added as metadata to the PDF, making specific information simpler to locate within it.

I prefer using Google Drive for collaborative work and sharing documents with others. This allows me to easily share and access documents from anywhere. For instance, my favourite recipes are online, easily accessible while shopping or to share with friends and family. Sharing and collaborating with my reading group members is done with Google Docs, for surveys and keeping notes of the books I've read and our recommendations.

Google Drive also has strong spreadsheet features for importing and analysing data. As I would expect, it also excels at searching for specific documents. Google Drive also has version control, which allows me to keep track of changes to documents over time. Its PDF management tools are also good. It is now far more straightforward to edit PDFs and add signatures if required. I have written a few scripts to automate the process of uploading documents to Google Drive.

However, I still prefer to keep my financial documents local for privacy and control reasons. Working with PDFs now is well supported by Linux, with tools like PDFtk, exiftool and LibreOffice, which allow me to easily edit and manipulate PDFs as needed.

rsync continues to be an essential tool for backing up my digital office. I have a regular backup schedule that ensures all my documents are safely stored.

Moving into the AI Era

AI is taking my digital office to the next level. I use Gemini not only for coding but also for managing my digital office. AI can help with tasks like searching through documents and extracting relevant information, making it possible to search multiple document types at once. I've only begun to explore the possibilities, but I can see how AI could help with tasks like summarising documents, extracting key information, and even generating content based on stored documents.

Google Drive is also getting an AI boost, providing better help and more advanced document interactions. There are many more features for me to explore here. Can I automate account reconciliation in GNUCash using AI? Can I use AI to analyse my financial data and provide insights? The possibilities are exciting, and I'm looking forward to exploring them further.