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...
Rehan ArshadTechnical Lead| Uxbert Labs / WebookRiyadh, Saudi Arabia
It is very important to know how the incoming data is stored, who has the read/write access, if there is a data lake, how data is transported to it and where is it stored, who can access the data lake, if and how this data is used to train AI models, who uses these models and who can fine-tune them. How is the consistency and integrity maintained? What level of redundancy is allowed and how the consistency is achieved? How long it takes to achieve consistency?
Having the data-flow diagram and list of the people with the access-type is very important. This list needs to be continuously reviewed and updated, because often employees leave the project or company and the still maintain the access which can cause serious data privacy issues.
Also, there must never be a single person responsible for making changes. There has to be someone to review and approve those changes. A second set of eyes will often save you from trouble. Saving Changes...
Narendrasinh RathodTechnical Manager| Value Chain Solution India Private LimitedIndia
In the development environment, training models often pose challenges due to insufficient data. Consequently, the accuracy of model results remains questionable. Additionally, collecting real-time production data in development or quality environments can be difficult. Saving Changes...
Daniel SapienzaeCommerce Manager - LAC| AnsellSao Paulo, SP, Brazil
Clear communication is crucial when dealing with unexpected data challenges. It's vital to be transparent with all stakeholders about the challenges you are facing and how they might affect the project. This approach will help build trust and ensure that everyone is on the same page. Apart from effective communication, it's also essential to be open to seeking help from external experts. If you're unsure about how to handle a particular data challenge, don't hesitate to reach out to a data scientist or other specialist. They can provide the expertise and guidance you need to overcome the challenge and get your project back on track. Finally, it's crucial to adopt a solution-oriented mindset when tackling unexpected data challenges. Avoid getting bogged down in the problem itself. Instead, focus on finding creative ways to solve the problem and move the project forward.
Saving Changes...
Mohun SundarSenior Program Manager| Health Care Service CorporationRichardson, Tx, United States
The main pitfalls that we can encounter when working with large amounts of data are the accuracy, relevancy, and appropriateness to align with business use cases. It's appropriate to use Data Governance protocols and standards to identify and correct these issues to protect our internal and external stakeholders from risk while still evolving our business model and way of working to add value to consumers' lives. Saving Changes...
As a project manager especially while handling sensitive projects, we are always on a lookout for data breaches and proactively monitor any abnormalities. Firstly, I assess the nature and scope of the challenge to understand its impact on the project objectives. Then, by collaborating with relevant stakeholders, data analysts and subject matter experts, we brainstorm potential solutions. Flexibility is key; we adjust project plans and resource allocation as necessary. Clear communication ensures everyone is informed of changes and aligned on the path forward. Regular monitoring and evaluation help in identifying further challenges early, enabling swift adaptation to keep the project on track. Saving Changes...
Debra HunterProject Management Consultant| Information Design Consultants, Inc.San Pedro, Ca, United States
Haven't done it yet. I am looking forward to figuring it out.
...
1 reply by George Bruton-Delaney
Mar 24, 2024 7:02 PM
George Bruton-Delaney
...
Hi Debra. While I haven't fully integrated AI yet, I've had productive discussions on this topic in my previous role. We focused on aligning AI with strategic pillars and ensuring mission and vision alignment. Currently, I'm exploring use cases for proof of concept in anticipation of future discussions.
Saving Changes...
Shayma IvankoProject Manager| High Tide Consulting Inc.Ajax, Ontario, Canada
The biggest challenge my organization has faced is dealing with large datasets provided by clients. We sometimes find issues with the quality and consistency of the data. Duplicates are always present and sometimes caught prior to data migration. In recent projects we've taken a step back to have the client reasses their data to ensure it is a clean data set. Saving Changes...
High-quality and abundant data is key for successful AI projects, including ensuring its variety and completeness. In my experience, data can present unexpected issues. To navigate these challenges, I focus on cleaning and refining the data to ensure its accuracy and relevance to the task at hand.
Saving Changes...
Deeksha AgarwalaSenior Quality Analyst| MRM WorldwideFarmington Hills, Mi, United States
Data Quality and Quantity. Staying current with security laws. Saving Changes...
Data quality can determine the outcome of your end product. Some times the volume of input data can make it difficult to deal with data quality issues. For example if dealing with 10 years of market data with millions of points it can be difficult to determine quality of the underlying data. In such cases use of AI can help classify the data and sift through it to determine outliers that can be reviewed for data quality issues. Saving Changes...
PMI Team Member
