Machine Learning Engineer (Image)

ANDPAD hires a machine learning engineer for image and drawing data from construction sites, building on AI features already in service such as drawing capture and blackboard creation. As site histories and drawings accumulate, the company wants to use them for new businesses and for improving clients' work.

The engineer runs proofs of concept for business problems, develops techniques to extract information from unstructured data such as drawings, builds products with multimodal LLMs and with machine learning, and builds pipelines that keep models running beyond the PoC. Depending on skills the role includes model serving, technical research, and technical presentations to stakeholders inside and outside the company, working with product managers and engineers. The stack includes Python, PyTorch, Vertex AI, MLflow, Streamlit, FastAPI, BigQuery and GitHub Actions, and the team defines the needed functions together with MLOps engineers.

Required are development and operation experience with machine learning, deep learning or LLMs, readable and portable code, setting metrics from business requirements, turning construction problems into solvable machine learning problems, and the ability to read research papers in machine learning, image processing or LLMs.