How might AI be applied to better understand the relationship between project risk and delivered value to stakeholders? One of my papers highlighted the importance of considering uncertainty, so I wonder if AI predictive analytics could improve modeling. Saving Changes...
Could you clarify the relationship you are looking at since delivered value is an "actual" whereas project risk is a "maybe". Are you referring to project risks which, if realized, could impact the expected value realized from a project?
AI tools might be able to identify potential value-impacting risks which have been missed but as far as predicting realized value outcomes, I'd rely on quantitative analysis tools such as Monte Carlo simulations to be able to provide confidence levels on value realization based on changes to key factors.
Kiron
...
1 reply by Claudio Szwarcfiter
Jan 27, 2024 12:19 PM
Claudio Szwarcfiter
...
Hi, Kiron. Yes, I'm referring to project risks which, if realized, could impact the expected value realized from a project. Now, in terms of value being "actual" as opposed to a "maybe", wouldn't you agree that in agile project management, value points are a maybe? It will depend on if we execute a certain user story.
Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
All related to risk can be supported by using AI. I am using it from years ago. Just to remember that: 1-AI is based on data. 2-You will obtain results (usually 3) with associated probabilities to each one. The human being decides. 3-Value is a different field. If you use AI trying to predict value is a different world to create the right algorithm.
...
1 reply by Claudio Szwarcfiter
Jan 27, 2024 12:23 PM
Claudio Szwarcfiter
...
Hi Sergio, thank you for your insight. You're correct, risk is more straightforward, value is something else. What about if the realization of a risk impacts my ability to deliver a certain feature in a product?
Could you clarify the relationship you are looking at since delivered value is an "actual" whereas project risk is a "maybe". Are you referring to project risks which, if realized, could impact the expected value realized from a project?
AI tools might be able to identify potential value-impacting risks which have been missed but as far as predicting realized value outcomes, I'd rely on quantitative analysis tools such as Monte Carlo simulations to be able to provide confidence levels on value realization based on changes to key factors.
Kiron
Hi, Kiron. Yes, I'm referring to project risks which, if realized, could impact the expected value realized from a project. Now, in terms of value being "actual" as opposed to a "maybe", wouldn't you agree that in agile project management, value points are a maybe? It will depend on if we execute a certain user story.
...
1 reply by Kiron Bondale
Jan 27, 2024 5:32 PM
Kiron Bondale
...
Claudio -
When I refer to "actual" value realized, I'm referring to a customer or other stakeholder using a delivered work item to solve their business problems or, in the case of a product development company, the ability to generate expected revenues based on a set of delivered work items.
However, going back to your original question regarding the application of AI to help, I will stick with my original response that Gen AI tools might help identify some benefits-impacting risks but likely greater benefit would be derived by using quantitative risk analysis tools to help quantify the degree of erosion of benefits caused by realization of some risks.
All related to risk can be supported by using AI. I am using it from years ago. Just to remember that: 1-AI is based on data. 2-You will obtain results (usually 3) with associated probabilities to each one. The human being decides. 3-Value is a different field. If you use AI trying to predict value is a different world to create the right algorithm.
Hi Sergio, thank you for your insight. You're correct, risk is more straightforward, value is something else. What about if the realization of a risk impacts my ability to deliver a certain feature in a product?
...
1 reply by Sergio Luis Conte
Jan 28, 2024 6:35 AM
Sergio Luis Conte
...
Meu Amigo. In my personal opinion and experience is the same than in other type of projects. Just to comment, about things are happend with GenAI in my actual work place, is GenAI (by the way, it is not new, it is outside there from 1979) can help to create some scenarios based on existing information. But something people miss about AI and GenAI in general: organization has to add the cost "to customize" GenAI actual foundation models to specific downstream tasks and organizations has to add the cost to perform outputs.check.br data-cke-eol="1" Unfortunately, as other things, GenAI is becoming a buzzword. (O Maradona é o maior que Pele...jeje. Abraco Argentino....)
In the product development world, it is common practice to estimate the market capitalization based on the provided capabilities and the price of the product, compared to the same price to value of competitors in that market space.
AI can certainly be used to apply variances to modeled variables to evaluate the impact of individual and multiple variables. Genetic algorithms can also be applied to evolve the model based on desired characteristics like resilience to the market factors responsible for many risks. Saving Changes...
Hi, Kiron. Yes, I'm referring to project risks which, if realized, could impact the expected value realized from a project. Now, in terms of value being "actual" as opposed to a "maybe", wouldn't you agree that in agile project management, value points are a maybe? It will depend on if we execute a certain user story.
Claudio -
When I refer to "actual" value realized, I'm referring to a customer or other stakeholder using a delivered work item to solve their business problems or, in the case of a product development company, the ability to generate expected revenues based on a set of delivered work items.
However, going back to your original question regarding the application of AI to help, I will stick with my original response that Gen AI tools might help identify some benefits-impacting risks but likely greater benefit would be derived by using quantitative risk analysis tools to help quantify the degree of erosion of benefits caused by realization of some risks.
Kiron Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
Jan 27, 2024 12:23 PM
Replying to Claudio Szwarcfiter
...
Hi Sergio, thank you for your insight. You're correct, risk is more straightforward, value is something else. What about if the realization of a risk impacts my ability to deliver a certain feature in a product?
Meu Amigo. In my personal opinion and experience is the same than in other type of projects. Just to comment, about things are happend with GenAI in my actual work place, is GenAI (by the way, it is not new, it is outside there from 1979) can help to create some scenarios based on existing information. But something people miss about AI and GenAI in general: organization has to add the cost "to customize" GenAI actual foundation models to specific downstream tasks and organizations has to add the cost to perform outputs.check.br data-cke-eol="1" Unfortunately, as other things, GenAI is becoming a buzzword. (O Maradona é o maior que Pele...jeje. Abraco Argentino....) Saving Changes...
George FreemanThought Leader | Author | Architect| Florida, United States
Hi Claudio,
Predictive analytics require vast amounts of contextually relevant historical data over lifecycle periods correlative to that used in your value projections. You need this data whether you are using a so-called “predictive AI application” or traditional predictive analytics in tools such as Excel.
In addition, if you expect quantitative outputs, we should recognize that “context is king.” Although we may want to believe otherwise, that context is only obtainable through human intelligence, especially in the fuzzy spectrum of soft value propositions, which often represent most of the value we see forecasted in non-commercial software projects.
So, suppose you have the contextually relevant historical data (which is the primary need). In that case, you are only a few steps away from applying the appropriate context, algorithm, and audit measures to each question to iterate an answer through traditional tooling. Versus doing this through so-called AI mechanics, where you would struggle with the modeling burden and reliability/trust concerns related to the outputs. Saving Changes...
