top of page
  • Writer's pictureAnantaya Pornwichianwong

A Day in the Life of Utt, Senior Machine Learning Engineer at Sertis

Machine Learning Engineer is one of the in-demand careers in the AI era, combining cutting-edge technology like machine learning with programming in the form of engineering.

A machine learning engineer might be a profession many are not yet familiar with compared to other AI-related jobs, but it is a crucial role that helps develop intelligent AI models that are behind various operations and drive progress in numerous industries.

Today, Sertis would like to introduce everyone to Utt, a Senior Machine Learning Engineer who is a key person in developing efficient AI models. What is Utt's daily work like? What tools does he use? And what aspects of the job make him passionate about this position? For anyone interested in pursuing a career in Machine Learning, this is a must-read.

What are the daily tasks of a machine learning engineer?

“Typically, a machine learning engineer's tasks involve working with models developed by the Data Science or AI Research teams and, if necessary, optimizing them for better performance. This optimization ensures the models are faster, use fewer resources, and are easier for users to utilize, whether by deploying them as APIs or creating pipelines. Optimization involves enhancing the models to run more efficiently and consume fewer resources. 

The machine learning engineering team also participates in technical discussions with customers. For some projects, we create APIs that customers can use directly, while for others, we develop services that the Software Engineering team will later integrate.

Moreover, the role of a machine learning engineer sometimes includes system design. When a client's requirements involve the model, the machine learning engineer is responsible for designing how the client will provide the data and defining each process. This task slightly overlaps with the data engineer's role. If the data volume is small, we handle it ourselves, but if it's large, a data engineer might create the pipeline, allowing us to focus on the model.

Additionally, during the design phase, we must ensure the infrastructure can support our model. We also discuss with the customer what the output should look like and how they will use it, incorporating business requirements into our planning.”

What are the daily tasks of a machine learning engineer?

  • 7:30 AM - 8:30 AM - Daily Exercise and Coffee I wake up around 7:30 AM, exercise until 8:30 AM, have a coffee, and then start work around 9:00 AM.

  • 9:00 AM - 12:00 PM - Project-based Stand-up Meeting and Tasks Since our work is project-based, we have catch-up meetings for each project. After the meetings, I review models from the Data Science or AI Research teams.

  • 1:30 PM - 6:00 PM - Reviewing Code and Implementing Services In the afternoons, I review code from team members and implement services to serve the models.

What are the typical tools a machine learning engineer uses?

Primarily, the tools used by a Machine Learning Engineer can be categorized into four main types:

  • Programming Languages: The main language we use is Python, and occasionally, we might use C++.

  • Machine Learning Frameworks: These include PyTorch, TensorFlow, and Scikit-Learn.

  • AI Development Platforms: We use various platforms depending on the provider. For the Google Cloud Platform, we use Vertex AI. For Microsoft Azure, we use Azure Machine Learning Studio. For AWS, we use SageMaker.

  • Optimization Tools include ONNX Simplifier, OpenVINO, and TensorRT

  • Cloud Services: These are tools we use when developing services and deploying them on the cloud. There are various platforms where we deploy our services, such as . For serverless services, we use Cloud Run to ensure users can access them efficiently.

What do you like the most about being a machine learning engineer?

"I enjoy working with data because I've always had a passion for it. Data holds immense power. Also, the role of a machine learning engineer doesn't involve as much mathematics as that of a data scientist, and since I also enjoy programming, I chose to focus on the engineering aspect.

Another thing I like about the role is that the work we do gets used by real users. It's gratifying to see our solutions being utilized directly by users."

How do you see yourself in the next 2 years?

"I still see myself as a machine learning engineer, but by then, I hope to have gained more knowledge. I aim to develop solutions that better meet customer needs, making models more efficient. Additionally, I aspire to share my knowledge with the next generation, helping them continue our work."

For more information about Sertis, including job opportunities, insights into our culture, and a glimpse into life at Sertis, visit our website:


bottom of page