Hey all,
I posted recently about finding a ‘system’ to manage my automations.
After thinking a fair bit about it I was able to come up with a V1.0 that so far is doing exactly what I wanted it to do.
Software I use (& used) to achieve this:
- Obsidian
- Integromat
- CODA
- Webooks / Mailhooks
I run most of my business out of these tools, multiple clients, multiple different systems, many automations… in short, it got messy trying to trace everything back to a source when implementing something new or sorting out an error. With a growing business, I needed to download this out of my head so someone can take over some of the ‘tech’ legwork.
Process
The idea was to have everything in obsidian with backlinks so I could see how everything fit together.
I achieved this by creating a naming convention in Integromat & Coda that allowed each system to ‘look each other up’.
Step 1 - was to rename all of my tables in CODA and ‘rename the modules’ in Integromat to match. I then renamed all of the ‘other modules’ (i.e. if a google sheet lookup was included, or a webhook or even a regex function). The result would look something like this [[CODA - ‘doc name’ - ‘table name’]]. I surrounded each of these with the double brackets so it would automatically create the note in Obsidian when imported.
Side note: any scenario’s that I need to update when on-boarding a new client I tagged with an emoji ‘outside’ the square brackets.
Step 2 - was to export the scenario’s from Integromat, which is fairly easily achieved by pushing a button. These then open for me in Visual Code Studio where I rename them (as the scenario name) and change the file from a JSON. to .md (Obsidian file). I then save these in a file in Obsidian.
Step 3 - In Integromat I created a scenario that runs periodically, looks up the list of tables in various documents and returns these. A lookup is performed to only return ‘new tables’. This is then exported (if there are updates), and added to Obsidian using the same process as above.
Step 4 - The notes are now in Obsidian and the graph populates immediately (the exact result I was after). I can now see which Integromat Scenario’s impact which CODA tables and what other data sets it talks to.
I also have another ‘Master Note’ in obsidian which queries the folder these Integromat notes and searches for the Emoji. This then returns me a list of Integromat Scenario’s I need to update if I onboard a new client.
That’s the basis of V1.0, so far it has made me feel much more ‘calm’ in the light of all my automations.
V2.0
V2.0 will involve adding in the additional ‘softwares’ that form part of these automations, including:
- Keyboard Maestro
- Shortcuts
- Various CRM’s
- Non Integromat API connections.
Plus I’ll also be looking into automating the 4 steps above.
I’ve already started looking into this using:
Hazel to grab the downloaded JSON file ➝ open it in Text Soap ➝ run a cleaner over the file and extract further data (for example, what rows are updated in a CODA table by the automation) ➝ then save it in the Obsidian vault file.
I’m learning Regex on the fly so it’s a ‘work in progress’, lol!
Any suggestions or questions I’d love to hear them!
Thanks all,
Danny
