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...
Akinwale AkinolaHead, Project Management| JNC International LtdSurulere, Lagos, Nigeria
Feb 27, 2024 12:47 AM
Replying to Gaurav Dhooper
...
Data quality, quantity, security and privacy have always been important and the best way is to prioritize them in any type of situation, project. They serve as the backbone of any organization's reputation and business.
I agree. Historically, while the value of data may not have been as apparent as it is today with the emergence of new technologies, issues related to data security—such as unauthorized access—have always been significant. However, these concerns have become even more critical now due to the immense value placed on data and the substantial consequences of a security breach. Saving Changes...
Mona FrazierHead of Change| Business Change Innovations, LLCNc, United States
Work hand in hand with security team and data scientists. Saving Changes...
I like the idea of prioritizing analysis. Saving Changes...
Jonathan LewisDirector of Software Engineering| IFG Companies Inc.Durham, Nc, United States
Cleaning data has always been an issue. Specifically, in research, identifying bad data, incomplete data, outliers, etc. has always been necessary before analyzing a data set.
The problem with GenAI is the type of data being analyzed. How do you "clean" qualitative "Big Data"? I don't know.
But, in ETL projects for example I have run into situations where 99% of data is transformed easily. However, the remaining 1% contains many unique data issues that cannot be transformed using the standard rules. My approach was more of a brute force method for identifying and addressing each type of unhandled variance from the standard. In the end the very small sets of variant data were manually transformed.
However, I do think an AI approach to the transformation should be possible. The high percentage of known transformations could have been used as training data then the smaller set variant data should be decipherable by basic pattern recognition. Saving Changes...
Data is a powerful tool, but it can also present pitfalls if not handled correctly. Here is an example of a challenge I've faced and how I navigated through:
In one of our construction projects, we relied heavily on historical data to forecast material costs and project timelines. However, we discovered midway that the data was incomplete and outdated, leading to inaccurate projections and budget overruns. To address this issue, I took the following steps:
- I conducted a thorough audit of the data to identify gaps and inaccuracies.
- I implemented a process for cross-verifying data with multiple sources to ensure accuracy. This included consulting with suppliers and using real-time market data.
- I worked with the IT team to upgrade our data management system, integrating it with live data feeds to ensure continuous updates and accuracy.
- I communicated the issue transparently with all stakeholders, adjusted expectations, and revised the project plan to account for the corrected data.
- I established stricter data governance protocols, including regular data audits and validation steps before using data in decision-making.
In this situation, the key was to stay proactive, communicate effectively, and implement processes that prevent similar issues in the future. This experience has reinforced the importance of robust data management and continuous learning in managing complex construction projects. Saving Changes...
Khaled El Haj IsmailHead of Programs| United Nations World Food Program UNWFPTripoli, Lebanon
Most of the projects I have managed with the United Nations relied on data collected in harsh conditions where technology is not advanced. The capacity level of data collectors was always a significant issue, and the constraints and risks associated with data integrity, confidentiality, and reliability were consistently high. Many project decisions depended heavily on the quality and timeliness of the data collected.
To overcome these challenges, I consistently relied on advanced data analysis techniques and collaborated with team members such as statisticians and data experts to enhance data quality. I prioritized data cleaning as a critical work package and ensured sufficient time was allocated for this process.
I worked closely with the team to conduct comprehensive risk assessments and develop a risk mitigation plan for all risks associated with data, along with several other techniques. Handling data in such contexts is crucial for project success but also incredibly stressful and challenging. I am genuinely eager to understand how AI can assist in improving data management in projects situated in harsh environments with low technological and digital advancement.
Saving Changes...
Leandro MartinsProject Manager| Klippel EnergyDomingos Martins, Es, Brazil
It realized during the development of a project that we were using a poor-quality data. It had to stop and built a task force to improve the process of gathering data throw their endpoint. Saving Changes...
George OffordPM II| FreelanceBudapest, Budapest, Hungary
Hi, While working many years in an enterprise telco environment the main issue has been to identify and retrieve relevant data from the humongous ocean of data that a telecom network generates. This task combined with the stringent requirements laid out by EU data compliance laws (GDPR) has made security and data retrieval (mostly in the form of KPIs) an enormously complex but interesting task. There is a wealth of useful information to be harvested but it is residing within an exceedingly large pool of data of which the bulk is not of use. Collecting all this data also has costs (HW), which makes realtime collection and disposal of irrelevant data a priority. This is a challenge. Saving Changes...
The primary issues I have encountered when working with large amounts of data are inaccuracies, unclear, and duplicate or old data. It's appropriate to use software to identify and correct these problems, and also closely monitor and synchronize the data, keeping database quality as perfect as it can be. Saving Changes...
PMI Team Member
