Please tell us about yourself. How did you end up in such an offbeat, unconventional and unusual career?
It’s somewhat fateful that Farshad Miraftab wound up attending graduate school only minutes from where he was raised. Growing up in Albany, California with parents who immigrated from Iran, Farshad knew he wanted to be an engineer — a particular field just didn’t come to mind right away. “In retrospect, thinking back, I maybe could have had more conviction,” Farshad laughed. His father, a structural engineer, ultimately inspired Farshad to pursue an undergraduate degree in Civil Engineering at UC Davis. It wasn’t until closer to his graduation in 2013 when his interests began to shift towards statistics and failure analysis, the process of using engineering principles and statistics to determine likelihoods of structural failure.
How did you end up at PG&E?
As a post-graduate, Farshad was out looking for work when he had a fateful encounter at Caffe Strada in Berkeley. He ran into a Risk Manager with Pacific Gas & Electric Company (PG&E) who worked in gas pipeline failure analysis as a team of one — and coincidentally was hiring. Farshad came on as an Operation Risk Analyst in August of 2013 as the team’s first hire and would continue on for the next year and a half. A lot of critical work was transpiring in this field; the tragic San Bruno gas pipeline explosion in 2010 prompted PG&E to incorporate a more rigorous data-driven decision making framework. This catalyst began to shape the regulatory landscape for all California utilities — how can data be used as a strategic asset to inform both operational and strategic business decisions?
Tell us about your work in data science
PG&E focuses on data-driven risk analysis to support safe and reliable gas and electric services to over 5 million households in Northern California.
Farshad helped in various roles while at PG&E, including Business Analyst, Senior Risk Analyst and ultimately a Data Scientist. He was responsible for building innovative probabilistic models that leverage asset data to quantify the risks on the pipeline system. He leveraged machine learning algorithms to provide intelligence in system failures including but not limited to, corrosion rates, inspection criteria, and equipment failures. However, as a Data Scientist, although its critical that you have the hard skills in building and implementing complex model, its just as critical that you can justify and explain them. Farshad’s role included presenting these models to PG&E’s executive team, lawyers, and even California regulators. “If you can’t explain a model to the right stakeholder, it’s meaningless,” Farshad commented. “We had to bring rigor to how decisions should be made: data-informed, as opposed to who’s just shouting the loudest in the room.Building and validating the model is one thing. Getting non-technical people to understand it is another thing.”
What did you do next?
Ultimately Farshad decided he wanted a formal education in data analytics. He felt the coursework and networking would complement his skillsets as a data scientist well. Upon researching UC Berkeley’s Master of Engineering program, he decided a degree in Industrial Engineering & Operations Research (IEOR) with an emphasis in Data-Driven Decision Analytics would well suit his career. His first class, Optimization Analytics with Professor Ilan Adler, was asserting and inspiring. Farshad recalled Prof. Adler writing a single sentence on the board: “Optimization is an art and a science.” He then turned to the class and said, “Notice that I put ‘art’ before ‘science.’” In the field, Farshad adopted this same philosophy. Classes like these made him appreciate science and engineering to an even greater degree. He was gaining better fundamentals in math and could leverage his coursework with PG&E as a data scientist. In August of 2017, Farshad transitioned to Asurion, a phone insurance and warranty company that also develops apps and products for customers to better assist their experience on mobile (connecting to wifi, managing photos, using Siri, etc.) He joined the ‘Big Data Analytics’ organization as a team lead for managing and improving customer engagement and growth on Asurion’s mobile applications. This entails using analytics to understand how customers engage with their products and creating models that can predict, and therefore reduce, customer churn.
Tell us about the course
Farshad, like any MEng student, shares high praise for the hybrid nature of the Masters of Engineering degree and how it plays into the engineering field. “A data scientist is like an internal consultant,” Farshad commented. “Soft skills are as important, if not more important, than hard skills.” The ideal engineer can communicate and present their information understandably as well as build and lead a team. Operating with an entrepreneurial mindset has been one of Farshad’s greatest professional assets.