InterMechanico
Well-Known Member
Thoughts on Constellation Software after the recent SAAS pullback? Probably different feelings for a holder vs looking to add
Hot stock tips
- Thread starter Rain City
- Start date
baddogg
Crew Member
It has been part of my portfolio for about a year. I wish I could say I was happy with it over that period.
ryanonthevedder
Well-Known Member
Everything is going AI, but don't buy the AI companies.
Buy the energy needed to run it all.
I am heavily invested in uranium. DMX.V if you like swinging at fences.
InterMechanico
Well-Known Member
I’ve added a few SAAS names to my portfolio since the large correction in that sector. CSU.to is one of them. I don’t think it will be an easy ride, and previously lofty valuations will probably not be revisited in the short term. CSU.to is on a very solid financial footing with a lot of cash to deploy but a changing landscape in which to work. People have long referred to CSU.to as BRK North, and the recent management shift to taking minority positions in larger companies is a step in the direction of that narrative.
Losing Mark Leonard as CEO is tough. He remains on as permanent adviser, but for some that won’t be enough to retain confidence. I look at this differently; Mark Leonard along with his management team built a very successful and productive M&A strategy over the past three decades or so. Te company is now evolving, not into a legacy/zombie company like IBM (I think people fear this), but into a company that deploys its assets into broader landscape that will still include small firm M&A but not exclusively. If shareholders still want the level of growth seen over the last decade, the company must adapt to its increasing scale to achieve this.
With regard to the omnipresent AI threat, it is there and it is real, but I feel it’s more likely that SAAS firms will be better positioned to make productive use of it than end users. Some end users will trim their seat counts to accommodate fewer employees being replaced by AI Agents, but for me the obvious end result is that seat cost goes up commensurate with productivity. In the short term, Agentic AI provides a loop hole for end users, but SAAS still provides the high value work flow tooling and will still get paid what they need for it, regardless of what form the end user takes.
Finally, as a small business owner and sole employee; I run a mobile Heavy Equipment repair business, I use Google Gemini Pro multiple times per day. I use it for investment analysis (not the paragraphs above!), tax planning and fact finding, procurement (super handy), and finally for troubleshooting. I have spent so many hours training Gemini to help build a bespoke troubleshooting assistant to voice ideas, find obscure technical bulletins, parse internet dialogue into actionable information… It has been a game changer. Having said that, I still need every legitimate subscription source of information I needed before. All the troubleshooting trees, detailed systems operation information, all of it. When I ask Gemini how a system works rather than “Summarize this System Operation Text”, this is where the hallucination (read: mistakes) happen. Great for the former, and can really help provide context, Not so great for the latter although it improves over time.
One last thing, (Sorry for the TLDR Post!)
Gemini suggested that I should think about moving to an on premises AI system to maximize u the utilization of data I already own, which in my case it’s approximately 2 TB of service information in various formats. This started a fun and informative, but ultimately unsuccessful journey to build a Local AI model. I bought a Mac Studio (M3 Ultra, 96 GB Ram), learned the basics (from Gemini) on how to use Python Code, downloaded Llama 4B, started indexing files, etc… 18 hours later, so much learned… Need at least 512 GB Ram, on and on and on.
The long and short of it, I still use Gemini Pro ($20/month), am waiting on availability of the next iteration of the Mac Studio M5 Ultra, I still use QuickBooks online and recently negotiated a 4 year agreement with my accountant (who is and will likely always be a human)
Sorry for the length of the post!
Losing Mark Leonard as CEO is tough. He remains on as permanent adviser, but for some that won’t be enough to retain confidence. I look at this differently; Mark Leonard along with his management team built a very successful and productive M&A strategy over the past three decades or so. Te company is now evolving, not into a legacy/zombie company like IBM (I think people fear this), but into a company that deploys its assets into broader landscape that will still include small firm M&A but not exclusively. If shareholders still want the level of growth seen over the last decade, the company must adapt to its increasing scale to achieve this.
With regard to the omnipresent AI threat, it is there and it is real, but I feel it’s more likely that SAAS firms will be better positioned to make productive use of it than end users. Some end users will trim their seat counts to accommodate fewer employees being replaced by AI Agents, but for me the obvious end result is that seat cost goes up commensurate with productivity. In the short term, Agentic AI provides a loop hole for end users, but SAAS still provides the high value work flow tooling and will still get paid what they need for it, regardless of what form the end user takes.
Finally, as a small business owner and sole employee; I run a mobile Heavy Equipment repair business, I use Google Gemini Pro multiple times per day. I use it for investment analysis (not the paragraphs above!), tax planning and fact finding, procurement (super handy), and finally for troubleshooting. I have spent so many hours training Gemini to help build a bespoke troubleshooting assistant to voice ideas, find obscure technical bulletins, parse internet dialogue into actionable information… It has been a game changer. Having said that, I still need every legitimate subscription source of information I needed before. All the troubleshooting trees, detailed systems operation information, all of it. When I ask Gemini how a system works rather than “Summarize this System Operation Text”, this is where the hallucination (read: mistakes) happen. Great for the former, and can really help provide context, Not so great for the latter although it improves over time.
One last thing, (Sorry for the TLDR Post!)
Gemini suggested that I should think about moving to an on premises AI system to maximize u the utilization of data I already own, which in my case it’s approximately 2 TB of service information in various formats. This started a fun and informative, but ultimately unsuccessful journey to build a Local AI model. I bought a Mac Studio (M3 Ultra, 96 GB Ram), learned the basics (from Gemini) on how to use Python Code, downloaded Llama 4B, started indexing files, etc… 18 hours later, so much learned… Need at least 512 GB Ram, on and on and on.
The long and short of it, I still use Gemini Pro ($20/month), am waiting on availability of the next iteration of the Mac Studio M5 Ultra, I still use QuickBooks online and recently negotiated a 4 year agreement with my accountant (who is and will likely always be a human)
Sorry for the length of the post!
wildmanyeah
Crew Member
More like I just recommend products that i
Get a cut back on and do nothing to improve your rate of return!
Bryan Allen
Well-Known Member
And do you ask about medical issues from the web instead of a professional health expert ????
MRWood
Well-Known Member
Yes, I do, as the web is the only one that will answer meAnd do you ask about medical issues from the web instead of a professional health expert ????
Last edited:
ryanonthevedder
Well-Known Member
I'd suspect most people here haven't had a used for ChatGPT or Grok or Gemini but try asking the AI for advice on a stock. It cuts the due diligence down quite a bit and has been a godsend for my portfolio.
Similar threads
