Great article, Maja! As you pointed out in the comments, those workflows could also have been done without AI, so I prefer to use models only when really needed (and minimise costs). I would be interested in learning more about the Engagement agent and whether you are not checking LinkedIn too frequently (twice an hour) and getting blocked at some point.
That is exactly right - you can tweak it to once a day if you want to. The point is that my daily LinkedIn routine (30-60 minutes on workdays) includes 10-20 comments to selected profiles during the posting period. I prefer to know precisely where Itoplace comments, rather than mindlessly scrolling through notifications or the newsfeed. I do like to capture some posts early on when the authors are still active and replying to comments to gain more visibility, but most of us are very predictable (when we post), so I aligned my posting time to theirs. Do what works for you <3
How long did it take you to get proficient with n8n? Any tips for ramping the n8n learning curve fast? I’m overwhelmed when I look at it and am referring to simpler tools like Relay.
Hey Ben, literally a week and an in-house developer for writing custom scripts. Funny that you mentioned- my first builds were made in the Relay app. I like it- it is super simple and it takes like 10 minutes to build and deploy an agent, but there are two reasons why we went with n8n. 1. Costs - Rely's free version is just enough to run like one agent at a smaller scale (I could do a weekly email with best preforming posts by selected influencers, but that was it) - actual problem: I did not understand how much a team of agents would cost me to run, and that is a big no-no for me. I have to be able to project the costs of deployment. 2. Number of steps - the workflows have to be more complex than those templates to get something great out of them. Not every step needs AI - it gets expensive - sometimes good old-fashioned workflows do the job quicker and cheaper. I like to use AI only when it really adds value, not just because I think it is cool (it is a challenge, but I am fighting the urge :)). My rule of thumb when it comes to build vs. buy - if this is one of the core business value creators and the existing tools are limiting, invest in exploring building your own thing, if it is just for practice/ no-code tools do the job just fine.
Sick agents there Maya, very inspiring and extra kudos for the in depth sharing. Rooting for you!
Great article, Maja! As you pointed out in the comments, those workflows could also have been done without AI, so I prefer to use models only when really needed (and minimise costs). I would be interested in learning more about the Engagement agent and whether you are not checking LinkedIn too frequently (twice an hour) and getting blocked at some point.
That is exactly right - you can tweak it to once a day if you want to. The point is that my daily LinkedIn routine (30-60 minutes on workdays) includes 10-20 comments to selected profiles during the posting period. I prefer to know precisely where Itoplace comments, rather than mindlessly scrolling through notifications or the newsfeed. I do like to capture some posts early on when the authors are still active and replying to comments to gain more visibility, but most of us are very predictable (when we post), so I aligned my posting time to theirs. Do what works for you <3
Careful with that LinkedIn scraping - playing with fire!
So true! I don’t have any automations connected to my profile - too big of a risk
Thanks for the in depth post.
How long did it take you to get proficient with n8n? Any tips for ramping the n8n learning curve fast? I’m overwhelmed when I look at it and am referring to simpler tools like Relay.
Hey Ben, literally a week and an in-house developer for writing custom scripts. Funny that you mentioned- my first builds were made in the Relay app. I like it- it is super simple and it takes like 10 minutes to build and deploy an agent, but there are two reasons why we went with n8n. 1. Costs - Rely's free version is just enough to run like one agent at a smaller scale (I could do a weekly email with best preforming posts by selected influencers, but that was it) - actual problem: I did not understand how much a team of agents would cost me to run, and that is a big no-no for me. I have to be able to project the costs of deployment. 2. Number of steps - the workflows have to be more complex than those templates to get something great out of them. Not every step needs AI - it gets expensive - sometimes good old-fashioned workflows do the job quicker and cheaper. I like to use AI only when it really adds value, not just because I think it is cool (it is a challenge, but I am fighting the urge :)). My rule of thumb when it comes to build vs. buy - if this is one of the core business value creators and the existing tools are limiting, invest in exploring building your own thing, if it is just for practice/ no-code tools do the job just fine.
This is awesome. Thanks so much for taking the time to reply here 🙏🏼
Love the detailed walk-through!
so glad you liked it - apprechiate the feedback
Happy Q4
Holy crap
haha- good or bad holy crap? :)))
Better than good!
so glad to hear that!
Thanks a lot & happy weekend
Hey Maja, how to get your Multi-Agent AI System for LinkedIn? Couldn't find a link in the article.