Hi, this section seems to be fairly quiet, the last post is from May, so I thought I'd share something.
I used Sierra Chart from 2009 to 2012, then took a long break until last year. Coming back was easy, because SC has hardly changed: same GUI, same studies. I still think it is one of the most reliable platforms for trading, and for me Denali is the best data feed.
In September I started using AI to build some minor studies. I was surprised what you can do with ACSIL / C++, especially with the GDI functions. Things I didn't think were possible. But I'm a trader, not an engineer, so it got messy. That's when the latest AI models, in my case Claude, turned out to be really good at helping me organize the project around one core study that everything else depends on, especially the GUI.
The result is my own interface inside SC: Numberbar in the chart, an order-flow DOM , volume profiles with a trade journal, and one Cockpit column for all settings. There's an interactive tour on Github: https://chrislb-trading.github.io/mts-tour/
Nothing is for sale. I asked the admins about sharing this on October 20 and haven't had a reply, so I'm posting it as a regular member. Admins, if this isn't OK here, please move or remove it. 
If there's real interest in a release later, I'll talk to you first.
SC may look old-fashioned, but there's a lot more in this platform than most people use. Has anyone else here built something with ACSIL and AI? What surprised you, and where did it fall apart?

chrislb 

asked the admins about sharing this on October 20 and haven't had a reply, so I'm posting it as a regular member.

Hi there chrislb

I don't recall any message about this. How did you send it?

In any case, as long as nothing is for sale, there's no problem.

Well hello there, so glad to find a fellow Sierra Charts Developer here. Isn't AI just amazing for building ACSIL studies, especially with GDI rendering. I use Gemini. I too find Sierra the best platform. I use a Denali data feed and an AMP feed for trading. Sierra's learning curve , if that's what you call it, was steep. The UI and documentation are just as convoluted, but Gemini has been a game changer. When you start diving into SC development you start seeing why things are the way they are with the UI, especially the study configurations and loading. I have built a number of studies. The development of SC studies still requires my years of software development expertise particularly when strange issues arise and AI responses start going in circles on. Its kind of like being an architect with a great junior developer who is very good at debugging complier errors and finding things/examples in documentation. But every now and then I need to dive into the code and back up on the direction AI has gone. AI can quickly go down a path that is way over course, overly complicated or just flat out wrong. Knowing when that occurs only comes from experience. An example of this is a response like "in this situation use a square wheel cause it will stay put better with no need for a brake". Do you make use of MBO data? It's the essence of the old saying "The more you know, The more you don't know". Take for example the timestamp on tics, it all depends on how you access them, how you want to process them and how you want to view them. Sierra is fast because of its exceptional architecture, but once you start doing things outside of that swim lane you need to pay close attention to the details which is many times in the what and how you access the data. The ability to make use of GDI is a game changer for the interface in the studies. I have often think perhaps I should provide some golden nuggets back to SC to help out their interface. But then I remind myself my focus is on being a successful trader. Looking forward to more conversations with you.