Practicum Pride: Manifold
Geoffrey came to the MSDS program after working for 4 years as a data analyst in Hong Kong and Shreejaya came to the MSDS program after working for 2 years as a business analyst at EXL in Bengaluru. Continue reading to learn more about their practicum experience at Manifold.ai!
Can you tell us a bit about Manifold AI?
Manifold.ai is an AI engineering company. Global companies partner with Manifold to build intelligent software products faster. Their Healthcare customers span pharma, med device, health IT, and payer. Their High Tech customers span electronics, networking, wireless, and streaming media. Manifold has offices based in greater Boston and San Francisco.
Manifold has a great team of engineers from leading tech companies, venture-backed startups, and Ph.D. programs. They have a very strong engineering and learning culture. It’s an ideal place where we can learn to solve the right problem and solve the problem right.
Can you describe the project(s) that you are working on?
We are working on a problem called system identification. The goal is to create a machine learning framework to infer the parameters of a dynamical system (system of differential equations).
Many physical phenomena can be modeled as dynamical systems such as fluid flow and chemical reactions. Normally we are able to model the structure and observe the outputs, but we don’t know the internal parameters behind that generate such outputs. Efficient and accurate parameter estimation algorithms can be applied to many industry problems. For example, we can estimate the impact of a specific drug on our body system parameters.
How are you applying the knowledge gained from the program to your practicum? Is there a particular class that has been the most helpful?
The first part of our project focuses on building a module for data simulation. Our coursework on data structures has been particularly helpful in the software development and design of this module.
As our project progresses to the parameter estimation module, our machine learning courses provide us with a good understanding of standard approaches and we know how to ask and consider the right questions when we approach new open-ended problems.
What is it like working with professional machine learning engineers?
Everyone has heard of buzzwords such as “big data” or “AI”, but most of the time the deliverables are in Jupyter notebook. The interesting part of the machine learning engineer role is that it is a mix of a software engineer and data scientist. They are responsible for both demonstrating the business applications of machine learning and delivering production software that customers can utilize.
We are grateful to have company mentors that instilled best practices around developing robust software since day one. The code review sessions provide us with opportunities to learn how experienced engineers analyze and organize code. We also learned some extremely useful software tools like Git, Unit Testing, CI, Docker, and MLFlow.
What is the biggest challenge you’ve faced in your practicum?
Not a lot of research that has been conducted on the problem we are working on (at least with our approach). Initially, we found it difficult to wrap our heads around this open-ended research question. However, with the help of our mentors we were able to formulate an approach and break the problems into workable chunks.
Find out more about USF’s MS in Data Science Program and their Practicum Program.
