Amanuel Teweldeberhan is a Senior Data Scientist at Blume Global, based out of the Pleasanton office. He brings great diverse experience and insights into the work he does at Blume. He was born and raised in Eritrea and his education and career have taken him to all corners of the world. In our Blumer spotlight interview with Amanuel, we explore his life experiences and learn more about what he does at Blume Global.
You got a PhD in Physics in Ireland! Tell us a little bit more about what made you get a PhD in Physics and what it was like living in Ireland for 4 years?
I had a passion for research-based problem solving in Physics and pursuing a PhD was the best option. I was also motivated by the research work I have performed during my MSc studies in Physics. Ireland is a great place to live and study. The people are friendly and sociable.
After getting your PhD, you ended up working for Livermore Lab for 4 years. What was that experience like? What did you do there?
I moved to Canada for a postdoc soon after completing my PhD and came to the US in 2009 to work at the Lawrence Livermore National Lab (LLNL). My research work at the Lab was studying properties of matter under extreme conditions using first-principles computational methods.
What made you interested in becoming a data scientist? How did your previous career experience help you in your current job?
The work I have performed at LLNL involved developing algorithms and running computer simulations which generated big data. Analyzing the big data to make insights required a lot of programming, data analysis, and data visualization skills. This helped me to make a transition to a career in Data Science.
What’s your current role at Blume and what are some of your key responsibilities?
I am working as Sr. Data Scientist at Blume Global. My responsibility is to build Machine Learning (ML) and Artificial Intelligence (AI) models by processing our historical data. The shipment and transactional data we have is vast and requires cleaning and transformation before using it for ML/AI. This involves large-scale data analysis and parallel computing for fast processing. We have to slice, clean, and merge the data utilizing computational resources such as Google Cloud Platform (GCP).
Large-scale data analysis… Can you expand on how you do that at Blume?
I have analyzed tens of terabytes data from simulations and IoT devices to build ML models before I started at Blume Global. One of the challenges we have at Blume is cleaning addresses written in different formats. I have developed a methodology which uses Natural Language Processing (NLP) and machine learning clustering techniques to clean and map the 1.2 million addresses in our database. The cleaned addresses will be used to identify if a new address written in any format belongs to our existing address database. This is also how we provide location awareness, enabling predictive ETA and accurate visibility.
Why did you choose Blume?
I have worked with various startups as a machine learning and data scientist after leaving LLNL. I became more interested in supply chain during my time at Bonsai AI. The goal of my project was to build an AI model which helps a supply chain analyst to identify the root causes of unfulfilled demand. This motivated me to join Blume Global which is solving interesting and challenging problems to disrupt the supply chain industry.
What excites you the most about working at Blume Global? Can you talk a little bit about the data science team?
Blume Global is building a unique platform powered by AI/ML to provide end-to-end actionable visibility of shipments by predicting vessel arrival date/time, port service time, and when a truck or rail should be booked to pick up a shipment. It is exciting to be part of the innovation.
The Data Science team has a big responsibility in accomplishing the goals of the company. The team is strong and is working hard to enhance the AI/ML capability of our platform.
What do you like to do in your free time outside of work at Blume?
I love to jog and spend time outdoors with my friends and family.