Global Data Science Forum

Data Science and AI from anywhere...

By Carlo Appugliese posted 10 days ago

  

By Carlo Appugliese, Program Director, IBM Data Science and AI Elite

blog_image.jpg

As we all learn about Social Distancing and transition to working from home, the Data Science and AI Elite (DSE) team is still open for business!

The DSE team is a team of the world’s top data scientists who work on use-cases to help IBM clients kick start or continue their journey to AI. With travel ban we’ve set up shop in our homes with our families, dogs, and other roommates splitting time in makeshift offices, including bedrooms, living rooms, closets and yes even laundry rooms. Who else can relate?

All jokes aside, first I want to highlight that Data Science and AI has many use cases which are very relevant to our clients in the current landscape. AI has the ability to provide a proactive approach to this crisis as clients try to address coming financial impact. For example, we recently helped a large bank create a model to predict loan defaults. This was focused on banks who supply their clients with both deposits and loans. The model helps predict the probability of a future loan default months in advance. This allows the bank to be proactive and offer programs that can mitigate default from occurring. The developed model has the ability to adapt to risk-adverse scenarios like the one we are living these days. This is just one simple example, but we have many AI use cases like this that we have developed with clients.

To continue our success with clients virtually, we have taken a 3 prong approach going 100% virtual. Starting today, we are offering:

 blog_image_2.png

1. Weekly Digital Broadcast

We launched a weekly digital broadcast, Chat with the Lab, which covers best practices around data science and AI, use case experiences, outcomes, case studies and any topic of interest collected from the team. Register and learn more here.

I recently was featured as a guest discussing Agile AI, and you can access the replay here.

2. One-on-One Briefings

A big part of what the DSE team does is meet with clients one-on-one and cover a variety of topics around data science and AI. To keep this going, we are running virtual one-on-one briefings with our clients discussing many topics around Data Science and AI, including our 4 part approach to the AI puzzle.

 blog_image_3.png

First part is anchored around a use case or business outcome, it important to make sure you don’t get lost in the technology. Next we bring the right skills and tools to solve each business problem. We then develop a working proof of concept using client data following an Agile AI methodology. This approach allows us to help clients quickly prove value with AI and reduce their risk of failure.

3. Engagement — Co-Creation of an AI POC

A successful engagement includes many one-on-one briefings, workshops and weeks of Agile AI code development as described above. To adhere to social distancing, we have gone 100% virtual so we can help our clients physically kick-start their AI projects with their data and our team without any risk of the virus.

 blog_image_4.png

Our approach has 3 parts, first is the Planning Phase where we shape a use case, looking into the data, review model approaches and defining a working team. Next we Stand up an AI Environment (Cloud Pak for Data) to do the work, grant access to the team and load the data. Lastly is when the magic occurs during the Prove Phase, we start sprint development and have daily stand ups and weekly sprint read outs. AI gives you the ability to address business problems before they happen and as companies look to face how current events will impact their business, this Agile AI approach gives them a working prototype in weeks vs months.

To learn more, visit the DSE Website to see how were working daily with our clients: community.ibm.com/DSE

Visit our Remote Data Science website for resources and best practices on how to conduct your data science projects remotely: http://ibm.co/RemoteDataScience

Download O’Reilly Agile AI e-book here: ibm.biz/agile-ai-ebook


#News-DS
#Highlights-home
0 comments
704 views

Permalink