Learning & Innovation Research Manager| Project Management Institute (PMI)Spain
Data quality and quantity is particularly important as we think about leveraging AI on projects. Considerations include the diversity and comprehensiveness of the data that is available to us.
Have you ever encountered unexpected challenges or pitfalls while using data in your projects? How did you navigate the situation and find a resolution? Saving Changes...
Katie DeBakeySales Operations, Business Analysis, Project Management, Data AnalyticsAustin, Texas, United States
I'd say that something to be aware of and to try to adhere to in using data from LLMs, is to have the LLM provide the sources is has pulled the data from so that it can be checked by a human.
Yes.. encountered a several challenges while working with data on projects.
To name one - Incomplete or inconsistent data - especially when sourcing data from legacy systems. In such cases, collaborated with data owners to understand the gaps and worked on root cause and corrective action to avoid such situations from recurring.
Saving Changes...
Juan Carlos Coronel LópezSenior Project Management| IDOM Consulting, Engineering and ArchitectureGetxo, PV, Spain
The data in organizations is an intangible asset that is rarely appreciated because its maintenance and standardization are not given priority within functional processes. Awareness campaigns and data purification and enrichment projects help to successfully achieve project objectives. Saving Changes...
I'd say that something to be aware of and to try to adhere to in using data from LLMs, is to have the LLM provide the sources is has pulled the data from so that it can be checked by a human.
Great help in reducing hallucination.. Saving Changes...
Sharad Kumar SaxenaEngineering Manager| GEDU Services Pvt LtdGhaziabad, Up, India
Jan 13, 2024 7:19 AM
Replying to Sergio Luis Conte
...
Key to understand is what you stated: "Data quality and quantity is particularly important as we think about leveraging AI on projects". This is key in AI from long time ago, from 1970s at least. But all related to data rest on other discipline that today is called Big Data to put it under and umbrella. And it is independent of you use AI or not. So, let me say, nothing new below the sun. Just to understand that data has to be converted into information. Again, nothing new. It was analized by Claude Shannon in the 1940s.
Very rightly said Luis, Data quality and quantity are critical considerations when leveraging AI in projects because they directly impact the accuracy, reliability, and overall success of AI models.
High-quality data ensures that the inputs to AI systems are clean, consistent, and free from errors or biases, which is essential for producing meaningful and unbiased outcomes.
Adequate quantity of data, on the other hand, provides the AI with enough examples to learn patterns, make predictions, and generalize effectively across various scenarios. Without sufficient high-quality data, AI models can underperform, leading to incorrect results, reduced efficiency, or failed implementations.
Thus, managing and optimizing data quality and quantity is foundational for maximizing the potential of AI in any project. Saving Changes...
Valerie BrownProgram Management| Anthem Inc.Fort Myers, FL, United States
Quickly identify and assess the nature of the challenge and assess the particular challenge's impact on the project. Saving Changes...
Challenges with data highlight the concept of garbage in, garbage out (GIGO): AI can process millions of data points, but it can’t work miracles if the foundation is unreliable. Volume is important, but quality is even more crucial. Have you ever faced a situation where you had plenty of data, but only a few were truly useful?
In a renewable energy project, we dealt with large volumes of hydrological data that were often biased. This directly impacted energy calculations, which were sometimes overestimated. The solution I proposed was to invest in refining the data set by combining data augmentation with validations conducted by external experts. It worked, but it shifted the project's focus: we began managing data quality as a critical step.
This makes me wonder: are we, as project managers, ready for this new reality? Refining data is now as important as scheduling or managing risks.
Saving Changes...
Scott C. PetersProject Manager | BriteSystems Northbrook, Il, United States
This is a great question, Claudia. On my platform migration and modernization project, we encountered pitfalls with data translations from EBCDIC to ASCII and how that was interpreted by the new system. The different encoding impacted the data’s indexing sort order, which changed the performance of the legacy system. It also had impacts on data imports from non-native systems and devices on the newer platform. All-in-all, we got things working through excessive hops and customization. Saving Changes...
Raman ChadhaManager| DeloitteMillbrae, United States
Maintaining high quality data is a shared responsibility, and needs to be imbibed into the DNA of any organization. I am currently working on an AI project that requires me to cluster projects. Upon doing exploratory analysis, I was left with only 25% data that I could actually use for the analysis. Just showing the quality of data to the stakeholders has triggered corrective actions, as people are sometimes unaware that they are relying on inaccurate data. Moreover, I like to show them (using synthetic data) the "what could have been" if we had good data - that is a very strong tool in nudging people to move towards maintaining high quality data. Saving Changes...
Thank you Nikita Jha for sharing this illustrative experience with data. Since we are encountering "new" problems we have to be creative in the solutions and yours seems to be a great option. I guess that finding the right resources for that fast-tracking approach was a challenge. Did you include specific data-related profiles? Have you detected any new roles needed in the context of project + data? thank you
Claudia,
I believe I believe that 'unexpected data' suggest that there is a break of pattern in data algorithm or progression. This can suggest that the new data encountered are outliers which upon further investigation reveal the reason behind the outlier. It may be worth your effort to find that reason in a means of understanding that data in particular. Additionally, in project management, risk, both foreseen and unforeseen, will give measures to take when unexpected data occurs. Root cause analysis may then increase the scope of previously determined risk. I find Ishikawa diagramming of 8P to a comprehensive enough to establish a good scope of risk. Lastly, I believe that the quality of the data is also established by the validation of the accuracy of the data tested. Depending on the type of data used there are several statistical test to do on both data instruments and data. Some of which are: Gage R&R, One way Anova, Standard Deviation Tests, Kruskal Wallis tests etc. It all depends on your data type and what exactly you want to determine. : Saving Changes...
PMI Team Member
