Woman of the Month März: Samantha Walker - Founder @mia
Heute wird euch Samantha von ihrem spannenden Werdegang erzählen. Sie ist Gründerin von mia, einer Plattform zum Entwickeln und Teilen von Machine Learning Applikationen.
From Building Buildings to Building AI
My name is Samantha Walker and I am the founder of mia, a no-code app builder for machine learning models.
I am originally from Canada, but have had the opportunity to live and work in both the US and Europe over the last several years. I started my career in the field of structural engineering after completing my undergraduate and graduate studies at McGill University. One aspect of the architecture, engineering and construction industry that frustrated me was how slow and inefficient it could be. I was always thinking of different ways that technology could be incorporated into our work to improve it and make it more efficient.
In 2015, I moved to San Francisco to work as a structural engineer for a large architecture firm. Being in the Bay Area, near Silicon Valley, fueled my interest in tech and sparked my interest specifically in machine learning and artificial intelligence. I thought, “If they can use machine learning to make self-driving cars, what might we be able to accomplish by applying it to the construction industry?”
In 2017, I traveled to Mexico City with a group of colleagues after a large earthquake to assist with post-disaster reconnaissance efforts. In the few days we were there, we collected hundreds of photographs of damaged buildings. We uploaded all of them to a dedicated portal to share our findings with the earthquake research community. For each photo, we needed to identify what type of damage was present and its severity level. Going through this painfully slow process manually, I realized it was the perfect opportunity for machine learning. I spearheaded an initiative to train a deep learning computer vision model to identify and classify damage from the set of building photographs we had collected. This experience led me to launch a machine learning focus group at the company, dedicated to exploring all the different ways we could integrate machine learning into our industry.
My experience leading this group is what ultimately inspired me to create mia. While we could build various machine learning models, our biggest challenge was being able to turn them into tools that others could use. It turns out this problem is widespread. Most machine learning models are stuck in code and never make it into production. This process is typically long, complex and highly technical. Mia seeks to break down these barriers by making it fast and easy for anyone to get their models out of notebooks and into tools that others can actually use. With mia, a machine learning app can be created in minutes by following a simple two-step process: (1) connect a machine learning model to mia and (2) create a front-end for the model through mia’s visual interface. Think form builder or website builder for your machine learning models. Our ultimate mission at mia is to make AI accessible to all by empowering data scientists to bring their models to life.
From building buildings to building AI, it has been an interesting journey and I am looking
forward to what the future holds for both me and mia.