Shayne PhillipsGlobal Executive| Dell TechnologiesQueensland, Australia
As we embark on our corporation's AI journey, we need to go beyond stakeholder communications and change management.
What are the key project management practices that we should prioritize and apply rigorously to ensure the success of our use case evaluation and selection process, given our limited budget? Saving Changes...
George FreemanThought Leader | Author | Architect| Florida, United States
Feb 20, 2024 2:04 AM
Replying to Shayne Phillips
...
Agree George, the Stakeholder engagement needs to be very precise and rooted in realistic outcomes. However, to return to the original question, we are with limited funds and seeking the must have PM capabilities for the "AI blueprint and adoption program". It's not an AGILE program and not fully waterfall, it will by iterative and use case driven, like an Innovation event.
As such Benefits measurement and realization projections will be key to make the work program a success and keep the EXCO investing.
Also, the multiple variations of strong risk management practices are an obvious choice. Specifically, DATA management and the watching over the associated cyber risks of that data getting outside of the business or into the competitors' hands.
Procurement, is it a make or buy situation, do we need to take a stack of funds and go to the open market, or should we build an internal Lab and test all this by for ourselves...?
Staffing/ Talent.. Hiring in the knowledge base, management consultants cost a lot, and this could be a multi-year program.
Yes, Scope control is a real heavy-duty practice to help keep everyone off the hyperbole!
Shayne,
Although I did not look deeply into this, I see that the organization you are with (per your profile) offers a “generative AI” solution/service.
Based on your statements/inquiries, it would appear that this solution/service resides outside your reach or does not fit your needs.
There seem to be some interesting corporate-political dynamics in play. If that’s true, those dynamics will likely dictate your approach—unfortunately.
It’s hard to make recommendations without additional context, but I’m a big fan of low-budget, small-team, hybrid “skunk works” projects. They can create explosive value when you have the correct talent engaged and the appropriate “covering” provided by executive management.
...
1 reply by Shayne Phillips
Feb 20, 2024 4:35 PM
Shayne Phillips
...
Hi George,
Correct, ultimately, some of our Gen Ai Solution / Services would be pulled thru on this engagement.
Establishing the core construct for the "AI Adoption program" is where we are today... but yes, an internal "SKUNK works" team of correct talent that is laser focused and using samples of core business data in a self-contained test LAB seems to be the best and most frugal option.
Proving the key use cases and then implementing at scale if approved.
The EXCO political dynamic is one of :- If successful I was involved, if not a success I was always skeptical.
So, with that context, the PM artifacts become key elements.
Based upon this chat and other conversations I am having, we need robust Risk management, a data breach is not an option. Scope/change and Procurement are the other two capabilities to spend upon early as must haves beyond Stakeholder engagement and communications.
Regular transparency reporting with progress gates should keep us in the correct frame / lane and on the correct path as we motor thru the use cases failing fast or succeeding soon.
Thanks for the chat it has cemented my thoughts..
Saving Changes...
Shayne PhillipsGlobal Executive| Dell TechnologiesQueensland, Australia
Feb 20, 2024 9:27 AM
Replying to George Freeman
...
Shayne,
Although I did not look deeply into this, I see that the organization you are with (per your profile) offers a “generative AI” solution/service.
Based on your statements/inquiries, it would appear that this solution/service resides outside your reach or does not fit your needs.
There seem to be some interesting corporate-political dynamics in play. If that’s true, those dynamics will likely dictate your approach—unfortunately.
It’s hard to make recommendations without additional context, but I’m a big fan of low-budget, small-team, hybrid “skunk works” projects. They can create explosive value when you have the correct talent engaged and the appropriate “covering” provided by executive management.
Hi George,
Correct, ultimately, some of our Gen Ai Solution / Services would be pulled thru on this engagement.
Establishing the core construct for the "AI Adoption program" is where we are today... but yes, an internal "SKUNK works" team of correct talent that is laser focused and using samples of core business data in a self-contained test LAB seems to be the best and most frugal option.
Proving the key use cases and then implementing at scale if approved.
The EXCO political dynamic is one of :- If successful I was involved, if not a success I was always skeptical.
So, with that context, the PM artifacts become key elements.
Based upon this chat and other conversations I am having, we need robust Risk management, a data breach is not an option. Scope/change and Procurement are the other two capabilities to spend upon early as must haves beyond Stakeholder engagement and communications.
Regular transparency reporting with progress gates should keep us in the correct frame / lane and on the correct path as we motor thru the use cases failing fast or succeeding soon.
Thanks for the chat it has cemented my thoughts..
Saving Changes...
Brian VickeryFounder and Principal Consultant| Actualize AdvantageRound Rock, Tx, United States
Hi Shayne,
Having only really ever been on and leading smaller teams with limited or non-existent budgets for implementing process updates I've found that it is best to rely on similar analysis that one would utilize to determine if any project should go forward; Business case/impact, cost/time to build, ROI, etc.
1. Identify target processes.
a. Look for pain points - Get feedback from the people doing the work!
b. What tasks or workflows are utilized the most?
c. Are there complicated workflows that are rarely used that slow users down due to unfamiliarity?
2. Analyze processes - Break down steps in a matrix, recording the task step, actions required, resource responsibility, tools required, time required, and dependencies.
3. Optimize processes to eliminate unnecessary steps and to ensure the process is still utilizing the latest best practices, tools, and the RIGHT resources.
4. look for opportunities to Automate steps and implement AI.
5. Evaluate opportunities for biggest impact and get to work.
Depending on what tools you are working with there can be concerns over data privacy and proprietary data, so it is important to ensure that there is corporate agreement on what data is "sensitive" and what is not. With that agreement in place a lot can be done to generate some of the project documentation in a way that it is robust enough to guide the project, but general enough that without additional documentation like specifications, drawings, schematics, etc. That it isn't a privacy concern to share with AI tools.
Best of luck, it sounds like you have an interesting journey ahead of you.
...
1 reply by Shayne Phillips
Mar 06, 2024 6:36 PM
Shayne Phillips
...
G'day Brian,
Spot on
" it is important to ensure that there is corporate agreement on what data is "sensitive" and what is not."
Data management, Access control, sensitivity, classification and gathering will be the at the epicenter.
Taking the risk out of this domain will advance the program significantly..
A risk management framework for Data seems to be the key artifact to position and start the journey.
A risk management program to provide adequate processes for identifying, evaluating, and treating risks around the organization’s valuable information.
It addresses uncertainties around those assets to ensure the desired business outcomes are achieved.
I see this as the key project management practice that we should prioritize and apply rigorously to ensure the success of our use case evaluation and selection process, given our limited budget.
Having worked on AI consulting assignments in my company, it is extremely important to define the scope of the project clearly. As technology is rapidly evolving, it can be tempting to jump on the bandwagon of the next new discovery.
Try to perfect one use case first and ensure that the project is completed within the stipulated time. As you have a limited budget, it is important to track cost of resources and time spent on implementation of the project. Rigorous testing also needs to be done both by QA personnel and business users as there are more chances of the output coming out wrong than right.
Hence, to summarize scope and cost management would be the most vital areas to track. Post realizing the benefits, you can scale the project. Saving Changes...
Navaid Ur RehmanAdditional Director / Project Management Expert /Writer /Trainer| Confidential (Pakistan)Karachi, Sd, Pakistan
Hi Shayne,
You can adopt Agile and Lean methodology.
Eliminate waste ,Eliminate non value added processes and activities,optimise resources and processes.Data management is also essential. Saving Changes...
William M Hayden JrAdjunct Assistant Professor| University at Buffalo, School of Management, Operations Management & StrategyBuffalo, Ny, United States
Given the track record of project failures for reasons unrelated to tech,
now with AI projects can fail faster!
Cheers,
Bill Saving Changes...
Shayne PhillipsGlobal Executive| Dell TechnologiesQueensland, Australia
Feb 23, 2024 10:29 AM
Replying to Brian Vickery
...
Hi Shayne,
Having only really ever been on and leading smaller teams with limited or non-existent budgets for implementing process updates I've found that it is best to rely on similar analysis that one would utilize to determine if any project should go forward; Business case/impact, cost/time to build, ROI, etc.
1. Identify target processes.
a. Look for pain points - Get feedback from the people doing the work!
b. What tasks or workflows are utilized the most?
c. Are there complicated workflows that are rarely used that slow users down due to unfamiliarity?
2. Analyze processes - Break down steps in a matrix, recording the task step, actions required, resource responsibility, tools required, time required, and dependencies.
3. Optimize processes to eliminate unnecessary steps and to ensure the process is still utilizing the latest best practices, tools, and the RIGHT resources.
4. look for opportunities to Automate steps and implement AI.
5. Evaluate opportunities for biggest impact and get to work.
Depending on what tools you are working with there can be concerns over data privacy and proprietary data, so it is important to ensure that there is corporate agreement on what data is "sensitive" and what is not. With that agreement in place a lot can be done to generate some of the project documentation in a way that it is robust enough to guide the project, but general enough that without additional documentation like specifications, drawings, schematics, etc. That it isn't a privacy concern to share with AI tools.
Best of luck, it sounds like you have an interesting journey ahead of you.
G'day Brian,
Spot on
" it is important to ensure that there is corporate agreement on what data is "sensitive" and what is not."
Data management, Access control, sensitivity, classification and gathering will be the at the epicenter.
Taking the risk out of this domain will advance the program significantly..
A risk management framework for Data seems to be the key artifact to position and start the journey.
A risk management program to provide adequate processes for identifying, evaluating, and treating risks around the organization’s valuable information.
It addresses uncertainties around those assets to ensure the desired business outcomes are achieved.
I see this as the key project management practice that we should prioritize and apply rigorously to ensure the success of our use case evaluation and selection process, given our limited budget.
Saving Changes...
Shayne PhillipsGlobal Executive| Dell TechnologiesQueensland, Australia
Thank you for exploring this topic with me.
" it is important to ensure that there is corporate agreement on what data is "sensitive" and what is not."
Data management, Access control, sensitivity, classification and gathering will be the at the epicenter of the work effort. Taking the risk out of this domain will advance the program significantly...
A risk management framework for Data seems to be the key artifact to position and start the journey.
A risk management program to provide adequate processes for identifying, evaluating, and treating risks around the organization’s valuable information.
It addresses uncertainties around those assets to ensure the desired business outcomes are achieved.
I see this as the key project management practice that we should prioritize and apply rigorously to ensure the success of our use case evaluation and selection process, given our limited budget.
