Merck Sharp & Dohme LLC - Senior Machine Learning Scientist
June 2024 - Present | Greater Philadelphia, PA
The Big Picture
Working at one of the world’s largest pharmaceutical companies has been like being a digital alchemist - except instead of turning lead into gold, I’m turning data into life-saving insights! My role involves deploying cutting-edge AI to make drug manufacturing more efficient, safer, and smarter.
What I Actually Do (The Fun Version)
Picture this: I spend my days teaching machines to see defects that human eyes might miss, building algorithms that can predict when manufacturing equipment needs maintenance before it breaks down, and creating computer vision systems that ensure every pill meets the highest quality standards. It’s like having a digital quality control superhero team!
The Tech Magic Behind the Scenes
My toolkit includes the heavy hitters of AI - TensorFlow, PyTorch, and Python for building the brains of our systems. I work with OpenCV to give machines superhuman vision capabilities, and leverage AWS and Google Cloud to deploy these solutions at scale. The coolest part? Watching a computer vision model catch a microscopic defect that could have affected thousands of patients.
The Impact That Matters
Every model I deploy, every algorithm I optimize, ultimately contributes to medications that help people around the world. There’s something deeply satisfying about knowing that the ML pipeline you built this morning might contribute to a treatment that saves someone’s life next year.
Cross-Functional Adventures
One of the best parts of this role is working with manufacturing engineers, quality control specialists, and process optimization experts. It’s like being part of a diverse superhero team where everyone brings their unique expertise to solve complex problems. I’ve learned as much about pharmaceutical manufacturing as I have about advanced AI techniques!
The Innovation Playground
Merck encourages innovation and experimentation, which means I get to explore the latest in ML research and apply it to real-world manufacturing challenges. From hyperparameter tuning on production models to implementing drift detection systems, every day brings new technical challenges that keep me on my toes.