A foundation in artificial intelligence, data science and machine learning
There is a lot of hype around artificial intelligence, data science and machine learning. While this may be the hottest technology discipline this decade, there continues to be a lot of questions about what it can do, where the latest technology advances are, and how to incorporate this capability into existing business practice. Our briefing will go in depth to provide practical advice on data science product development, current technology, approach to talent, and data delivery needs of most organizations. You will leave with a perspective on what cultural changes, technology changes, and talent is needed to create and maintain a thriving data science capability.
Data science 101
Both data and science are required for a viable practice to become a reality. Most organizations believe they are swimming with data but have limited ability to mine for and gather insights that can produce significant business value. We will examine the shift in perspective from lakes and warehouses to pipelines that enable machine learning. We will provide an overview of what machine learning is and how to differentiate hype from reality when reviewing technical literature. We will also cover the scientific side to expose bias, the scientific method, and the significant milestone of models in production.
Machine learning frameworks and architectures
There are no standards for developing, deploying, and managing machine learning in production. Academic and entry level data science is usually performed on a local environment such as a laptop as this is where every data scientist begins their journey. Local environments do not scale to production applications and are severely limited in their ability to examine medium and large-scale datasets. Current technology presents many options for doing data science and subsequently deploying models to production applications. Our overview will tour a typical journey from local to batch predictions to cloud analytics to streaming models to edge deployment of machine learning.
Data science product development overview
Some organizations originate data science efforts from IT. This can lead to misconceptions and missteps when searching for engagement and opportunity from business areas. The process for creating data science products is different from IT projects whether using a waterfall or scrum methodology. Agile practices are effective in data science projects, however, the steps are distinctly different and the risks of projects should be managed accordingly. We will cover the CRISP-DM process as well as the primary risks of data science projects.
Use case discovery
Machine learning projects are in fact data products and benefit from a product owner and management perspective when initiating and framing use cases for data science. We have developed a method for discovering and framing data science opportunities rooted in product development. We will run through this discovery process with an example that exposes your team to the types of questions relevant to product development.
* Disclaimer: This offer is subject to qualification, no substitutions and no rain checks issued. Not valid toward previous purchases. Other restrictions may apply.
A representative will reach out to you as soon as possible to get you started on your journey.
We look forward to helping you implement new, positive initiatives in your company.