Hi Charlie, please tell us about your journey from being the Manufacturing Engineering Manager to becoming the Senior Director of Product Management, AIML, and Cognitive Analytics at Oracle?
In the 1980’s I was a Manufacturing Engineering Manager at Automatix, Inc., a pioneering start-up adaptive intelligence, robotics, and machine vision company. Previously, I had graduated from a three-year Manufacturing Management Program at Honeywell Computers that involved time from work to study for dual graduate degrees in Manufacturing Engineering and an MBA both at Boston University.
My planned career path was to gain knowledge and expertise in manufacturing, especially state-of-the-art manufacturing, and pivot that expertise into a software company. My first foray into Software was as a Product Manager for the Operations Advisor, at Palladian an expert systems company in Cambridge, MA, with close ties to leading MIT and Harvard faculty in this space. All the other product managers and LISP developers at Palladian were M.S. and Ph.D. academics who knew theory. Still, none had ever worked in a manufacturing setting, so my knowledge of the target user served me well. Later, I joined Bolt, Beranek, and Newman, Inc., applying my manufacturing domain knowledge as a Product Manager for manufacturing SPC/SQC, statistical data analysis, and experimental design software, so again my manufacturing domain knowledge served me well in that product management role.
It both has and has not changed much. I still meet with developers and customers for various machine learning new features, use cases, and predictive applications, but today I use Zoom. As much of my role involves developing working relationships with others, I almost always have my camera.
The acceleration of everything: new information, new ideas, the development of new products and features, customer adoption and customer feedback, the success or failure of a product in the marketplace, and if successful, the quickness of competitors to respond.
The rise of data and how it is leveraged to make the customer experience better. Not so long ago, doctors kept your medical records in manilla folders. Now, customers and patients select with whom they choose to work within no small part on the business’s or physician’s perceived ability to know everything about the customer or patient and have 100% accurate, secure information and be able to act on it immediately. Today, you can compete just on the product, quality, and price. Today, business and organizations have to know their customers and patients, accurately remember everything about them and anticipate their needs and their preference on what and when to engage in relationships with that customer or patient.
Yes, that was big. On Dec. 5th, 2019, Oracle communicated in a blog post their decision to make Oracle Machine Learning (OML) as well as Oracle Spatial and Graph no license required features of Oracle Databases. Previously both had been priced Database Options, but when they were included in the Oracle Database license, customers really noticed. Now, in the case of Oracle Machine Learning, customers can perform machine learning inside the Database eliminating the architectural, data movement, data security, and incompatible platform challenges associated with using traditional, dedicated machine learning platforms, e.g., Python, R, SAS, IBM/SPSS, Data Robot, etc.
Now with Oracle, the data never leaves the Database. All machine learning model building, model application, and deployment of OML’s new insights and predictions can emanate from the same industry-leading data platform. This changes everything. Instead of cycles of weeks and sometimes months to build but more importantly deploy machine learning models, Oracle’s strategy to “move the algorithms, not data” collapses that cycle to minutes or days. OML model building takes full advantage of the Database’s parallelism and scalability while preserving user access and data security. OML models can be used to “score” new data in the same Database, moved to another database production and transaction processing, or called from REST services.
Honestly, to me, it is the industry’s awaking to the power inherent in a next-generation hybrid data management and machine learning platform. There is no reason that enterprises can’t start thinking about data management of historical data and data management of “predictive” information. Oracle has been doing this for over a decade now, but with Oracle Machine Learning’s library of 30+ in-database implementations of ML algorithms that are exposed via a range of languages and user interfaces, SQL, Python, R, REST, “drag and drop” U.I., notebooks and embedded in applications. Data of all forms, structured, unstructured, transactional, spatial, graph, sentiment, and images can be combined to harvest more information for immediately improved customer relationships.
I enjoy reading a lot of science fiction and books about how machine learning will change the future and most times think, “Exactly! I can’t wait for that new capability to be available.”.
Additionally, I always try to learn. Knowledge is the key. Take time to learn how to do something new, learn a new skill, a new machine learning language, a new technique, new Software, or just the detailed answer to a customer’s question. Anything you learn combines with your previously accumulated skills and knowledge to give you an advantage.
Start now. Do something. Don’t wait. I do not recommend going outside and hiring a Ph.D. Data scientist. Instead, grow your citizen data scientists and data scientists from within. Leverage the people who already know the business, understand the data and can get things done. Hire consultants or trainers who can “coach up” your data analysts to become more knowledgeable in machine learning concepts and the application of machine learning algorithms. Formulating a well-defined business statement, assembling the “right data,” understanding the data, and developing clever “engineered features” that extract more domain knowledge from your data are more critical than which algorithm to use. At Oracle, we make machine learning simple with automated data preparation, AutoML, and even auto data insights. Machine learning is about automatically extracting information and discovering insights. Start now to learn to take advantage of machine learning for competitive advantage and then start embedding ML models into every application possible.