McAfee’s Celeste Fralick: “I’d like to see greater synergy between the seemingly endless amount of cybersecurity companies; If we work together, we are so much more powerful against the adversary”
I’d like to see greater synergy between the seemingly endless amount of cybersecurity companies. If we work together, we are so much more powerful against the adversary. As in adversarial machine learning, we, as researchers, must be much more proactive in predicting what the bad actor can and will do, and implement corrective action prior to catastrophe. Like chess, anticipating your opponent is critical and eventually allows you to win.
As part of my series about the women leading the Artificial Intelligence industry, I had the pleasure of interviewing Celeste Fralick, a Senior Principal Engineer and Chief Data Scientist at McAfee. Dr. Fralick is responsible for McAfee’s technical analytic strategy that integrates into McAfee consumer and enterprise products as well as internal Business Intelligence. She brings over 36 years of industry experience to McAfee — prior to Intel’s divestiture of McAfee, she was Chief Data Scientist in Intel’s Internet of Things Group where she developed Machine Learning and Deep Learning analytics for over 8 different markets.
Thank you so much for doing this with us! Can you share with us the ‘backstory” of how you decided to pursue this career path?
My career started as a Quality Engineer with Texas Instruments almost 40 years ago! My first assignment was implementing Statistical Process Control, a manufacturing data science technique that had been utilized in Japan in the 1950’s. This career kind of pursued me! As my career advanced through Fairchild Semiconductor, Medtronic, and Intel, I found myself gravitating to manipulating data, statistics, and other techniques in every one of the positions I held.Working on my Master’s and Ph.D. in Biomedical Engineering focused my efforts in advanced statistics, neural networks, and artificial intelligence — fortunately, right before the Machine Learning hype came along.
Sadly, cancer struck and, while I was recovering, I was presented an opportunity to lead data science for Intel Security — which just happened to be the previous McAfee. I returned to work learning about cybersecurity and being Chief Data Scientist for the now-spun-out McAfee.
While I remain cancer-free, I still undergo chemo every three weeks while working. I suspect the doctors are tired of me bringing in charts and graphs and lecturing them about cybersecurity! It is very humbling to realize I have been a data scientist all throughout my career!
What lessons can others learn from your story?
I am a strong supporter of focusing on your strengths rather than your weaknesses. Data is my passion, but it took a long time for me to realize it — it really was staring me in the face for a long time! Perseverance and tenacity have helped to focus on the things I enjoyed; having cancer only amplified that enjoyment of doing something I love.
Can you tell our readers about the most interesting projects you are working on now?
It’s always fascinating to work in this field! One is improving our data science fundamentals and ensuring that all the data scientists are doing the model right (“verification”) and doing the right model (“validation”)!
Additionally, my team is focused on attacking, detecting, and defending adversarial machine learning where, with minimal perturbation, the expected malware can be ignored! Scary stuff. It is, however, one of the first times that researchers are actually doing the attacking — not the bad actor yet — and finding ways of detecting and defending against these types of attacks. Typically, our industry has been very reactive where the bad actors have concocted a new threat and we respond. While that’s not likely to go away, adversarial machine learning gives us a glimpse into being much more anticipatory about what attackers can do.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
I have had great mentors along the way, as well as managers that have allowed me to dip my toes into areas that may not have fallen perfectly into the planning process. With almost 40 years of experience, I can sadly say I have had only three great, out-of-this-world managers. It really underscores “you work for the person, not the company” and has helped me be a better leader, learning from the best and ensuring you don’t repeat the mistakes of a bad example.
What are the 5 things that most excite you about the AI industry? Why?
1. Endless possibilities. We have not even begun to scratch the surface, particularly in the cybersecurity industry, and it’s really the art of the possibility at this point. Game theory is exciting and really in its infancy in this industry.
2. Technical nuances. The cybersecurity industry’s characteristics can be as unique as the human body — everything has to work and be optimized as best as possible to keep the bad stuff (or bad actors) out! The technical nuances provide interesting challenges in the feature and design space for productized solutions.
3. Passion in new data scientists. While we obviously need MANY more women in cybersecurity, many new data scientists are filling that gender gap and are very passionate about what they do. The zealousness is catching and invigorating to everyone around!
4. Teaching. Data science has been viewed as that black magic that nobody seems to understand. Audiences are anxious to learn, and it provides data scientists the opportunity to share our knowledge in a way that can be understood by even non-technical people. At the same time, we as data scientists must speak in the language of the CEO.
5. Growth. Growth for our companies, our industries, our people, and our intellect. It’s an exciting time to be part of this technology, even to see it applied in other areas of our lives! We’re definitely not standing still.
What are the 5 things that concern you about the AI industry? Why?
1. The hype. It is a roller-coaster of hyperbole. It challenges data scientists in a negative way; we have to stop and address the hype, taking us away from developing models.
2. The “fake data scientists”. While we don’t want to throw the baby out with the bath water, it takes time to teach others not formally trained how to do data science correctly. As typing then coding used to be foundational, it would be great to see the basics of data science incorporated into K-12 curriculum; analytics should be the new foundation (especially with the projected growth of data).
3. Lack of a statistical foundation. What most “fake data scientists” lack is the statistical foundation to understand their data prior to applying an analytic model. Even many university programs don’t include a foundation in statistics before throwing code into the mix. Taking the time to understand the math is critical, even before tackling simple code such as Python or R.
4. Lack of post-production analytic reviews. While the development of models is exciting, following up on how the model continues to work in the field is not. How the model “decays” due to change at the customer’s implementation point, how often the model “learns”, or even how to minimize false positives or false negatives can all impact a company’s bottom line. More emphasis needs to be placed after the model is in use.
5. Elegant solutions when easy can suffice. It’s like wearing a tuxedo to the beach — it serves the purpose but you’re definitely over-dressed. AI can be very marketable but not everything has to be solved with a fancy solution. Statistics have been around at least since the mid-17th century. Sometimes shorts and a t-shirt serve the purpose — and cost the company much less — than a fancy tuxedo.
As you know, there is an ongoing debate between prominent scientists, (personified as a debate between Elon Musk and Mark Zuckerberg,) about whether advanced AI has the future potential to pose a danger to humanity. What is your position about this?
With all great discoveries come unintended consequences. Our RSA keynote actually addresses one: Orville Wright was concerned about how planes could be used to wreak havoc; dropping bombs on innocent civilians is definitely horrific and unintended consequence of his discovery. The web has brought knowledge to our fingertips, but human smuggling and pornography can victimize the innocent. Ethical standards as well as human-machine teaming are great opportunities to enable AI integrity. Human machine teaming allows the low level and redundant tasks to be left to machines, while humans can continue to drive higher level tasks. As factory automation displaced some factory workers, those same workers embraced even new technologies, skills, and competencies. Should we have stopped the revolution of air travel, the web, or even factory automation? As humans, we adapt and evolve. We will do the same with artificial intelligence.
What can be done to prevent such concerns from materializing? And what can be done to assure the public that there is nothing to be concerned about?
Checks and balances via international standards are helpful, but not a panacea. Self-governance is not enough, and regulation is often over-reaching; a healthy balance is called for with compromise. Scientific development of AI solutions within ethical boundaries is often what is required to keep these concerns in check. Case in point is the topic of unconscious bias that the AI community recognizes within model development and is implementing new methods, corrections, and communications to combat. Is it enough, and will there be instances where the concern will materialize? Probably not, and likely so, respectively. We need to be vigilante, demonstrating ethical science, to assure the public. And we need to do this consistently, not just a “one time” thing.
How have you used your success to bring goodness to the world? Can you share a story?
I’d like to think so. Each model I develop prevents a bad actor from impacting another customer. Previous models that I’ve developed demonstrate that if you reach across industries, you can develop promising new areas. My Ph.D. dissertation predicted when an elder would have an emphysema attack in the future. While basic research, I’d like to hope that in both healthcare and cybersecurity, I have laid the groundwork for others to follow and improve our lives even further.
As you know, there are not that many women in your industry. Can you share 3 things that you would you advise to other women in the AI space to thrive?
Frankly, I suspect these are for any women — or men — in any space, but long gone is the “work-life balance” of the 1990s. Our lives are much more integrated now due to 24/7 communication.
1. Do not be intimidated. Do what you believe and follow your passion, despite the roadblocks whether they be related to your gender or not. Use your emotional intelligence to bridge gaps.
2. Don’t stress. It can cause health challenges not only now but in the future. Be a duck and let the worries roll off your back like water! Set boundaries for yourself and for your family. What are your priorities in life?
3. You don’t have to be perfect to be happy. Being positive and not-so-perfect is much easier than being negative and stressed. It will help in #1 and #2 above.
Can you advise what is needed to engage more women into the AI industry?
We, as employers, need to look to other degrees in addition to computer engineering. For example, Biomedical Engineering classes are typically 50% female and yet are poorly recruited. These students have a passion to create for the greater good, are trained in human-machine interfaces, and many have aligned with other departments, including data science, math, and statistics. As potential employees, we must look at our strengths and not be encumbered by our degree and typical recruiting companies. You never know what might happen!
What is your favorite “Life Lesson Quote”? Can you share a story of how that had relevance to your own life?
Change what you can, ignore what you can’t, and have the wisdom to know the difference. Wisdom comes with great patience, and often the slow drip-drip-drip of change drives you bonkers, but you will be stronger as your master the journey of patience. As a cancer survivor (with lifelong chemo), happiness and patience comes from strange bedfellows! I am much happier post-cancer diagnosis — I find I am ecstatic to wake up every day and feel the sun on my face, a family who loves me, and work I enjoy. What more can I ask for?
You are a person of great influence. If you could start a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. :-)
Wow, great question! I am flattered that you think I am a person of great influence! I have five to start!
1. I’d like to see greater synergy between the seemingly endless amount of cybersecurity companies. If we work together, we are so much more powerful against the adversary. As in adversarial machine learning, we, as researchers, must be much more proactive in predicting what the bad actor can and will do, and implement corrective action prior to catastrophe. Like chess, anticipating your opponent is critical and eventually allows you to win.
2. I envision the integration of our medical data to predict disease and to minimize waste as well as cost. Insurance companies must release their hold on medical data (without the PII, of course!) to allow crowd-sourcing for data scientists for the greater good. We have some of that in the cybersecurity industry, but we could do so much more to thwart both disease and the bad actors.
3. My home town of Lubbock, TX — as well as many Hometowns — has a struggling downtown. I’d like to see industries, entrepreneurs, academia, and non-profits utilize data, infrastructure, and economic acumen to drive the re-birth of dying and blighted downtowns. I realize this isn’t AI-centric, but there’s no reason why we can’t utilize AI to decrease poverty, increase economic growth, and enable our infrastructures to re-enliven downtowns.
4. There’s got to be less false-positives and false-negatives as well as increased comfort for that mammogram. It saved my life, but we’ve got to make it less painful and more precise.
5. Lastly, as a recent visitor to Antarctica, I have witnessed the pristine of the untouched world. We have destroyed our Earth, and we as its stewards need to own that accountability. What can AI do to make that happen faster? Clearly, our current solutions aren’t the entire answer but perhaps Artificial Intelligence will be more intelligent than what we have been thus far.
How can our readers follow you on social media?
https://www.linkedin.com/in/celeste-fralick/
This was very inspiring. Thank you so much for joining us!
About the Author:
Tyler Gallagher is the CEO and Founder of Regal Assets, a “Bitcoin IRA” company. Regal Assets is an international alternative assets firm with offices in the United States, Canada, London and United Arab Emirates focused on helping private and institutional wealth procure alternative assets for their investment portfolios. Regal Assets is an Inc. 500 company and has been featured in many publications such as Forbes, Bloomberg, Market Watch and Reuters. With offices in multiple countries, Regal Assets is uniquely positioned as an international leader in the alternative assets industry and was awarded the first ever crypto-commodities license by the DMCC in late 2017. Regal Assets is currently the only firm in the world that holds a license to legally buy and sell cryptos within the Middle East and works closely with the DMCC to help evolve and grow the understanding and application of blockchain technology. Prior to founding Regal Assets, Tyler worked for a Microsoft startup led by legendary tech giant Karl Jacob who was an executive at Microsoft, and an original Facebook board member.