Tokio Marine HCC

The beginning of the ITIL 5 journey has commenced. PeopleCert Verified Digital Badges - ITIL Foundation (Version 5) https://lnkd.in/eNbBjWTf

Joining the meme.

Day 8 – Practicing SQL Date Functions ⏳ Today I practiced the TIMESTAMPDIFF() function in SQL and explored how date-based calculations are used in real-world analysis. Worked on queries related to: • Employee age calculation • Experience calculation • Working days analysis • Resignation duration • Average experience analysis One thing I understood clearly today is that date functions become much easier once the logic between “start date” and “end date” is clear. Instead of only memorizing syntax, I tried focusing more on understanding the business meaning behind each query. 💡 Key takeaway: TIMESTAMPDIFF() is not just a date function — it helps measure the time gap between two business events. Sharing a small practice recording of the queries and outputs. #SQL #SQLLearning #DataAnalytics #LearningByDoing #TIMESTAMPDIFF

Every "AI caused this" story I've read this year has the same shape: 1. Something breaks. 2. AI was somewhere in the room. 3. Therefore AI broke it. The argument only works if you've never looked at what was happening before AI showed up. Uptime Institute has tracked outages for years. ~40% of organisations had a major outage from human error in the past three years, with 58% caused by staff skipping procedures. Power and networking lead the causes. Not AI. Verizon's DBIR has put the human element at 60 to 82% of breaches every year for over a decade. Credentials and phishing, not algorithms. Tech layoffs? 2001 cut 168,000 jobs. 2008 cut more. 2023 was the worst year for tech cuts since the dot-com crash, most of it before any serious AI deployment. The failure modes are old. The costume is new. Blaming AI is just a way to avoid admitting the problem was always us.

I really enjoy working in DevOps and most of all I love exposing new teams to what is possible with AI. Over the last two placements I managed to create meaningful impact and here is how I did it. Mindset before I joined: "AI is just a tool and you can ask it questions. Like a better form of google really ... " Mindset after I joined: "AI is how you should be orchestrating your work, forget about editors, it is a complete paradigm shift and I will never work in the same way again ... " The interesting thing people see, is the speed at which things happen in my world. Usually at some point they ask me in a 1:1 setting: "How the hell are you doing that?". My answer is simple, I explain that I setup my access to literally everything via the terminal, then I orchestrate each and every one of my actions through an AI console interface beit copilot, claude or gemini. I do not drag tickets from planned to in-progress, I get the AI to do it for me. It follows a process, a series of events which becomes part of a bigger workflow, checking code out, exacting the change, creating a branch, raising a PR, monitoring that all tests pass and monitoring for approvals. Once the approval arrives, merge the PR, monitor the deploy, then close the ticket with an update in the comments of what happened. All without so much as even browsing a single website. So that is the paradigm shift I can see here, the one I teach fellow collegues. It also surprises me how many people are still on the MCP track for building up their workflows using AI, this is simply not true. You can use the API directly without even knowing how it works, AI does that for you. When I show this to new-adopters, it is almost certainly followed by a second question: "Do you think we will be without a job in 6 months?". My answer is almost certainly always no. It is just changing the our way of working, nothing else. We are just moving faster, the old days of having to browse 10 different UI's is most certainly over. AI has killed the UI to some extent is all, not your job.