Senior Machine Learning Engineer - Predictive Modeling

GCS Recruitment; Baltimore (MD); On-site; USD 140,000 - 190,000; Full time

Job summary

GCS Recruitment is seeking a Senior Machine Learning Engineer to join a team building predictive models for construction and operations. You’ll own production ML models, analyze performance, and engineer features from historical and geographic data to improve accuracy.

The role emphasizes retraining pipelines, exploring new predictive problems, and collaborating with engineers to productionize ML solutions in an AWS-based environment with modern tooling.

Qualifications

• Strong ML engineering background with production experience.
• Experience building predictive models and improving accuracy.
• Proficient in Python and ML tooling (MLflow, DVC).

Responsibilities

• Own and enhance production machine learning models.
• Analyze model performance and identify areas for improvement.
• Develop feature engineering strategies using historical/geographic data.
• Build automated retraining pipelines.
• Develop predictive models using algorithms such as XGBoost.
• Work with time-based and historical trends to improve predictions.
• Evaluate factors like county, ZIP code, job type, seasonality, holidays, and processing times.
• Develop models for construction duration and permitting.
• Contribute to predictive cost modeling with historical labor/data.
• Explore ML approaches for automating construction design.
• Collaborate with engineers to productionize ML solutions.

Skills

Python

XGBoost

ML pipelines

AWS

Feature engineering

ML engineering

ML tooling

Tools

MLflow

DVC

HashiCorp Nomad

FastAPI

Job description

Senior Machine Learning Engineer - Predictive Modeling

We're expanding an ML team focused on applying machine learning to real-world construction and operational problems.

We're looking for a Senior Machine Learning Engineer who can step into an existing production environment, understand how the models work, identify areas for improvement, and help build new predictive capabilities.

The team currently has a production model that predicts construction job duration and is looking to strengthen that model while developing additional models around permits, costs, and automated construction design.

What Makes This Role Interesting?

You won't be starting from a blank page.

There is already a production model performing well across approximately 80% of use cases. Your challenge will be understanding the remaining 20%:

• Why are those predictions inaccurate?
• Is additional feature engineering needed?
• Are there specific use cases that require a specialized model?
• What additional historical signals could improve prediction accuracy?
• How can the model be retrained and kept current?

You'll also have the opportunity to work on new predictive AI problems.

What You'll Do

• Own and enhance production machine learning models.
• Analyze model performance and identify areas for improvement.
• Develop feature engineering strategies using historical and geographic data.
• Build automated retraining pipelines.
• Develop predictive models using algorithms such as XGBoost.
• Work with time-based and historical trends to improve predictions.
• Evaluate factors such as county, ZIP code, job type, seasonality, holidays, and historical processing times.
• Develop models for construction duration and permitting.
• Contribute to predictive cost modeling using historical labor and material data.
• Explore ML approaches for automating construction design.
• Work with engineers and the broader team to productionize ML solutions.

What You Bring

• Strong Machine Learning Engineering background.
• Experience developing predictive models.
• Strong Python skills.
• Experience with XGBoost or boosting algorithms.
• Feature engineering experience.
• Experience analyzing model performance and improving accuracy.
• Experience building ML pipelines and production systems.
• AWS experience.
• Experience with MLflow, DVC, or similar ML tooling.
• Strong understanding of supervised machine learning.

Team & Technology

The current team includes ML resources in both the U.S. and India and is continuing to grow. The environment is AWS-based and uses HashiCorp Nomad, FastAPI, DVC, MLflow, and Python.

If you're an ML Engineer who likes digging into data, figuring out why a model isn't working, and turning those findings into better production models, we'd love to hear from you.

GCS is acting as an Employment Business in relation to this vacancy.