🎇Decision Intelligence Architecture enables organizations to gradually and incrementally achieve higher degrees of intelligence that improve and influence business operations and change organizational behaviors.

Yesterday, I covered how to achieve up to the “Intelligence Core.” based on #DecisionIntelligence Architecture. Read more detail at https://lnkd.in/gJGfiacE.
Hence, this layer of intelligence leverages symbolic AI and is deterministic, so organizations can confidently deploy these models safely and confidently to the most critical business operations. Not only that, but these models can also become the core for interoperability and standardization.

If you think about it, “Intelligence Core” is like a brain transplant: building intelligence that can be hosted inside any application, process, system and workflow.

🤔So what's the caveat?

The main challenge for this layer of intelligence is the lack of connectivity to outside environments. This means every consumer should pass the required information to these models, and then they will process the information and return the outcomes. But what if consumers of the brain do not have access to all the required information? After all, much of the information is related to or supported by operation systems, processes, and databases that consumers generally do not have any clue about.

Additionally, the result of decision execution in some complex yet common scenarios may not necessarily be one result or answer. It may, in fact, be a set of best choices and options for a specific situation. What if we needed the model to optimize the result based on some constraints or expected circumstances?

💡This is when the next layer of intelligence is needed, the “Expert Intelligence”. To achieve the intelligence of this layer, Expert Intelligence, the decision models require:
– #Optimization to resolve decision conflicts and competing priorities and understand goals
– Advanced #data connectivity, on-the-fly data processing so the decision can reach out to systems and processes and create information as needed based on situations and contexts
– Long-running #orchestration and human #intervention so decisions can interact with internal/external resources, systems, and processes and involve human actors to override decision outcomes or enhance inputs as needed

🚀Therefore, your models at the “Expert Intelligence” can interact with live data feeds and data sources, process them, plan and reason, solve puzzles, and optimize the results based on constraints and goals. At this layer of intelligence, the outcomes are still all transparent and explainable. They are accurate and deterministic based on domain knowledge and achieved “Intelligence Core”.

Tomorrow, I will explain how to progress toward the next two layers of intelligence…

Posted here.

Published June 4th, 2024 at 07:30 am