Take the AI Expert Challenge ‌‌🤖‌

[Michal_Gebacki]

Welcome back, Dynatrace Community!

Only three years ago many of us heard about AI occasionally at the utmost. Now it's everywhere. As with most of breakthrough technologies, if used wisely, it can help us rethink our daily routine. When cannon and dynamite emerged, you needed to be really careful to use them in the rightful way. Thankfully, with AI it's a little bit easier 😉

Share with us your examples of smart and creative usage of AI tools. How is it helping you? What struggles did it help you to overcome? Did something actually improve because of AI in your life? Hit us with your tips, maybe it will encourage others to automate and avoid some daily life struggles in easy way! 👇

Benefits of taking Community Challenges!

👉 Every participant receives a unique "AI Expert Badge"

👉 You will also get +100 bonus points for extra activity

👉 Engage with others and have fun!

[AntonPineiro]

Hi,

I am not using it in my private life, but for work, I use it for speed up some automations scripts mainly.

I had just tried about Dynatrace questions, but it is not correct, normally, and I still prefer documentation or this forum 

[MaximilianoML]

You know, I tried the same asking about Dynatrace and ChatGPT 5 can't easily answer correct, so my workaround was to "Teach" him with the Dynatrace Docs, I just attached every doc link, so he could store on his "memory", and then for my surprise the technical questions had a significantly result. But he still messing with innovating DQL logics, and I understand because sometimes is confused 🤣

[AntonioSousa]

Because Dynatrace moves so fast, I really can't get good answers from the AIs I use..

But it's gotten way better. Some year ago, it was just allucinating all the time. Now it's getting way better!

[MaximilianoML]

This is such a cool topic!

I'm a Tony Stark fan since I was a kid, so control stuff with simple voice orders is the coolest think I recently focused to achieve. I bought an Alexa, but she's limited because doesn't really use AI to power her answers and actions. So, I found ways to provide "my own AI assistant" to work with her. I linked with Claude Sonnet 4 and something called TriggerCMD, so now when I need something "better" from her I make her ask to "Jarvis" and then she returns his answer or even actions. It's so cool, but without this TriggerCMD Tool this should not work for now 😅

[AntonioSousa]

@Michal_Gebacki , this is an excellent and fascinating challenge!

My history of hate about AI started in 1992 with Prolog. Couldn't stand it! 

It got much better in my first Dynatrace Partner Summit, way back in 2016. At lunch, someone we all know, gave me one of the best replies in my life... The question: When will the Rise of the Machines happen?

4 years later, it became real when I met Davis in person at Perform!

And than, in December 2022, all changed! I started making some non-professional questions to ChatGPT. And in early 2023, two events changed everything for me:

1. My first WOW moment was in late January 2023. It was a challenge from a client that had a very good algorithm for detecting when a disk would get full. I had discussed this extensively in a lot of places, namely here.
   On 25/01/2023, I finally got the spec of the algorithm from the client:
   

   I looked at this and tried to get help from Google. I forwarded this to other people, even inside Dynatrace. I tried everything to understand it. I even created a spreadsheet and saw that it seemed to give good results. But I couldn't find out what this was.

   Three days later, on a Saturday, I did this:
   

   It struck me hard! Hours and hours of effort, multiple people contacted, and this was so easy? So, I made a decisive question, one for what I do not have a screenshot of the time, but that I have managed to find in ChatGPTs history:

   
2. In April 2023 was developing a Dynatrace extension that involved parallelizing network requests at a very high pace. I was asking ChatGPT for the code (as I didn't know how to parallelize code at the time in Python) but it wasn't working as intended. I was seeing clearly it not working with tcpdump traffic. But ChatGPT was insisting the code was OK. I then asked:
   

   I then gave it the output of tcpdump. Seconds later, I got a new version of the code, which I immediately checked and found working!

I have not looked back since then!

[ShriPSingh]

Open to ideas here. I am working on extending observability include AI models.

[Patrick_H]

I use it mainly to get new inputs on problems I try to solve. e.g. once we had some parts in our selfmonitoring logs that looked like garbage displaying random characters, as if there was a memory issue on some devices. But as it somehow looked like it followed a pattern I asked langdock about it if those characters make sense, and surprise - it just were numbers in a format I did not know up until then (perso-arabic numerals). So problem solved an learned something new.

[O11y_Noob]

every time i ask MS co-pilot / Chatgpt to answer anything related to DQL related to metrics , it answers start with Fetch <metricname> , now i have a DQL cheatsheet and attach to the conversation everytime i ask and also mention fetch will not work on metrics 🙂 it work very well for Dynatrace when we quote the doc and then ask solutions from it

[p_devulapalli]

That's a nice idea, I'll give it a try 😀

[tracegazer]

AI has basically become my unofficial co-worker — the one that never sleeps, never complains, and somehow answers faster than anyone else.

I use it for a bunch of things:

• Automated scripting: Turning vague ideas like “I need a script that fixes my mess” into working Bash/Python/PowerShell/K8s YAML.
• Documentation & communication: It helps me write messages that sound way more organized than what I could produce at 1 a.m.
• Prototyping & troubleshooting: When I hit a wall, AI helps me validate logic, test approaches, and remember commands I definitely should know.
• Personal life: I also use AI to analyze patterns in stocks and cryptocurrencies. Nothing crazy — just identifying trends, risks, and behaviors. And honestly… it’s been working surprisingly well.

Overall, AI helps me automate the boring stuff, avoid rework, and communicate clearly without overthinking everything.

And yes - I know there are AIs that can already make coffee.
But if we dream bigger… AI could help us solve problems that really matter: healthcare, education, climate resilience.
If it can help clean up my scripts, maybe it can help clean up the world too.

[dannemca]

MS Copilot is my best friend now... 

I understand programing language, of course, but aint not a developer.

Before co-pilot, I spent hours in stackoverflow looking for pieces of codes to implement my automations, mostly using Dynatrace API for reporting and bulk configuration.

Now, I just ask and got the py code done, quick and clean.

In personal life, I am using to edit photos (years of photoshop training is now retired) and start thinking in use it to help me with my finance guidance (stocks analysis, what and when to buy).

[DanielS]

Very similar use here!!!

[DanielS]

I have a dietary restriction, and AI has been a huge help in managing it. It helped me create a personalized meal calendar with recipes I can cook either in a traditional way or using a Thermomix as an alternative. In the latter case, it even provides me with the exact programs I should use! This has made my daily cooking much easier and more enjoyable.

[p_devulapalli]

One of the first use cases I've used AI for is for writing DQL back when it was newly introduced and it was also one of the first lessons for me on not to trust everting AI says . It use to literally create functions and commands that never existed in DQL and too many hallucinations.😄

But once I started understanding DQL and writing my own queries  , I've been using AI to improve the query performance. It does provide good suggestions based on DQL best practices in improving the query performance and smarter ways of formatting the query.

Its been a good leaning experience , but with a bit of caution 

[Malaik]

Hi,

Im using the AI to generate some usable piece of code:
Python, Mashup and ..(for powerBI)...

and daily, just to generate some funny photos and ask regular generic questions....

[Georgi_Vuldzhev]

I’ve been experimenting with using AI to make my workflow smoother and more organized. Using Ollama together with an Anthropic model, I built a small local knowledge base that I can query with natural, human-language prompts. I fed the system with my own documentation, notes, and references, and now it returns accurate, context-aware answers whenever I need them. The idea is to test if this would be a possible solution for everyday questions that I get from colleagues, but founding the answers based on company-specific data and not the publicly accessible ones.

Curently I am testing it on a local ProxMox homelab and it lacks resources, so it's terribly slow but I think it might be something worth playing with.

[gschramm]

I using ChatGPT to help me manage my homelab which is built with Proxmox. I'm running bunch of LXC containers and VMs together with an external NAS for storage. I'm not really experience in Linux that's where ChatGPT is helping me out a lot (troubleshooting, analysing logs...).. very handy for me to get prebuilt linux commands that just need to be executed.

I've also used it to help me plan our family vaction last summer (roadtrip with kids). I almost felt like cheating.. instead of endless hours of googling for suitable accomodations and activities it provided a tour within a few minutes.

Another usecase is product research / comparision. Whenever I'm about to buy something important I'm using it to provide in depth analysis and compare the different options.

Where it really helps me is getting started with something. When there is new challenge / task ahead and I don't know how to approach it AI is helping to get over that initial mental block fast. It's a kickstart tool for me.

[Mike_L]

Except for the constant addition of LLM results when using Google, my main usage is to create quick scaffolding for a script. When generating too much code it takes almost as long to understand (and fix) what it did, as it would have taken to create it myself, but it's great with creating the basic code to then build upon.

[PacoPorro]

drafting emails, summarizing content, explaining concepts, this is my main usage.
Even creating tidal playlist 🙂 

[ChadTurner]

AI can be great. I've used it to add clarity on things. For example, if I felt like I've gone too far into the weeds, I'll ask AI to "Rewrite for clarity" or even as we build out our Jira, I'll ask AI to "Rewrite this to follow Jira best practices". So it takes my ideas/thoughts and refines them for a clearer understanding. 

I've also used AI heavily in the research segment. Research this error message, or even tell me about a certain EXE especially if the SME of the application in question is Out of Office. Keep in mind, I always call out that I am not the SME and that I'm merely just using AI to give me a basic understanding of the entity at hand, that I would always recommend syncing up with the SME. 

AI has been great, is it perfect? No, not at all, but it does help. I just always double check it and remove anything that is not incorrect or misleading. 

[Kenny_Gillette]

Just like @PacoPorro 

Rewrite emails/texts/messages, understanding different items, overview of different subjects.