During my work as a business/data analyst, I have found a way to organize my project notes, scripts, dashboards and spreadsheets.
PART I - Paper notes (Yes I still keep paper notes)
1 - Prepare a batch of 10-20 prints of my special note-taking pdf for each project, depending on the size (small if within 4 sprints, and big if more than 4);
2 - Write down names of attendees during the meetings, so that I know who to contact when I need info
3 - Prepare a folder for each project and always put the notes, one page for a meeting, in the following order:
Top -> Always the most urgent or important meeting notes. E.g. KPI meetings, or telemetry meetings
Then I put the rest meeting notes on reverse chronological order (most recent at top)
In this way, I can always review the previous meetings, as well as the most urgent issues in under 5 minutes, which is good because I work in Agile and in most of the companies that I worked for, this means Fast, not Efficient.
PART II - Electronic notes, scripts, dashboards and spreadsheets
I have a top-level One-Drive folder called "Company_Team_Name", which is also shared between Win10 (Host) and Ubuntu (VirtualBox). I have two child folders under it, "Company_Team_Name/Project" and "Company_Team_Name/Task".
I use the following tools:
Excel, for results of analysis and other stuffs;
SQL, my bread and butter;
Python, sometimes I write Python scripts to help my analysis. I do not like notebooks so I write fully operational programs;
Shell, some Linux command line tools, like csvsql, are very useful for simple tasks;
Tableau, for all dashboards;
Typora, for all research note taking
I organize the project folders in Datagrip, as I pretty much need SQL for every report.
1 - In Ubuntu/Win10, create a new folder under "Company_Team_Name/Project" with the name of the new project, assuming it's a 4+ JIRA long project, not an overnight task.
2 - In Datagrip, create the following folders for each report (I worked long enough to know that we need these reports):
"./Pre_Analysis" -> For pre analysis of the project, usually some data dumps or small analysis of similar features in the past;
"./AB_Test";
"./Small_Analysis" -> We usually release a feature to only 5%-20% of the population for technical sanity check and small analysis (actually same as full analysis but only on most important KPIs);
"./Full_Analysis" -> 100% population Full Analysis, usually performed twice because CXOs will require a more detailed look after reading the first full analysis;
"./Tableau" -> put all dashboards here
3 - SQL scripts, Python scripts, Typora mds and Excel spreadsheets will be grouped under each report sub folder
In the meantime, I'm considering writing a shell script, or a Python script to automate creating those folders. I'm also considering using each Project folder as the venv directory for Python, but since all projects use the same Python package configuration (pandas, odbc, etc.) I just install packages on the primary Python interpreter.
This is a configuration that I find to be extremely efficient to work in an Agile environment.
