Tile Logo for Apps with AI features

[alexatwoodhead]

AI is fundamental to emerging software. The Open Exchange should display an AI feature logo on the Application tiles.

[evshvarov]

Great idea, Alex! do you know how to distinguish AI enabled startups automatically?

On Thu, Jul 27, 2023 at 7:26 PM Alex Woodhead @.***> wrote:

AI is fundamental to emerging software. The Open Exchange should display an AI feature logo on the Application tiles. [image: image] https://user-images.githubusercontent.com/691812/256630423-798c83fa-9241-4c38-b4b2-4baa8f00ab50.png

— Reply to this email directly, view it on GitHub https://github.com/intersystems-community/openexchange/issues/274, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAVHEP7C5YY42GLFHDJBERTXSKXGZANCNFSM6AAAAAA22PADZU . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[alexatwoodhead]

I think a scan of python classmethods and python source code for "import" and "from" statements, that utilize a known list of AI specific module names could be an approach. Additionally this will help projects that include python classmethods, yet don't get a "Python" logo.

[Lena-Ev]

We discussed this technical possibility and realised the data fetched maybe not accurate and missleading. So we suggest to ad another checkmark on app registration: my app utilises AI or ML / or two separate checkmarks.

[alexatwoodhead]

Interesting there is opportunity of more than one checkbox: A few years ago we may have looked at:

• "AI" is an everything superset that includes "ML".
• "ML" contains "DeepLearning" Therefore checking "AI" would also check "ML", so not that useful. Last year might have envisioned functional scenarios:
• AI Model Training
• AI Inference - Classificaton, Transcribe
• AI Generative - Transformer services like LangChain, CoPilot, ChatGPT Looking at other vendor as-a-service product formulation, can see a pattern that follows more a "Usecase" scenario. If there were only two checkboxes I would suggest for Open-Exchange:
• IRIS ML - Focused on "Smart Database Tables". Solutions with examples of implementing and configuring classes / tables to be trained by core IRIS product workflow
• Bring-Your-Own ML - Which is focused on developing external Training, Inference and Generative services that are interoperated or otherwise contribute to a solution with IRIS. For example LangChain.

[Lena-Ev]

@alexatwoodhead thank you! That looks more like dropdown list, but no problem to implement, will do!

[Lena-Ev]

Implemented as tags and filters, as well as the logos for AI and ML