Machine Learning Engineer - Production ML/AI

GCS Recruitment; Cherry Hill Township (NJ); On-site; USD 120,000 - 160,000; Full time

Job summary

GCS Recruitment is seeking a Machine Learning Engineer to support and expand a production ML environment for construction operations. You will understand existing models, improve accuracy, build data/retraining pipelines, and help move predictive solutions into production.

You'll work on problems such as construction duration, permit, and cost prediction, with an emphasis on production ML, MLOps, and AWS. Strong Python and experience with XGBoost or LightGBM are preferred.

Qualifications

• Experience in machine learning engineering and predictive modeling.
• Strong Python skills.
• Experience with XGBoost or LightGBM.
• Experience building ML/data pipelines and production ML/MLOps.
• Experience with AWS and experiment tracking (MLflow).

Responsibilities

• Analyze existing production models and identify improvements.
• Build automated model retraining pipelines.
• Perform feature engineering and analyze historical data.
• Investigate model errors and improve underperforming scenarios.
• Develop predictive models from structured/historical data.
• Deploy ML applications into production environments.
• Track experiments and model performance.
• Build data pipelines for training and feature engineering.

Skills

Machine Learning Engineering

Predictive modeling

Python

XGBoost/LightGBM

Feature engineering

Model evaluation

ML pipelines

Production ML / MLOps

AWS

MLflow

Tools

DVC

FastAPI

HashiCorp Nomad

Job description

Machine Learning Engineer - Production ML/AI

We're hiring a Machine Learning Engineer to help support and expand a growing production ML environment focused on Comcast's construction operations.

This role is ideal for someone who enjoys more than just building models. You'll be responsible for understanding existing models, improving their accuracy, building data/retraining pipelines, and helping move predictive solutions into production.

The team currently has a production model that predicts how long construction jobs will take and is looking to improve the model while expanding into additional predictive use cases.

The Opportunity

You'll work on several machine learning problems, including:

Construction Duration Prediction
Improve an existing production model that predicts how long construction work will take.

Permit Prediction
Develop a model to predict how long it may take to obtain government permits required for construction.

Cost Prediction
Build a predictive model using historical material and labor expenses to estimate the cost of future construction work.

Automated Construction Design
Explore ML approaches that could help automate aspects of construction design, including trench placement, poles, and cable layouts.

Your Responsibilities

• Analyze existing production models and identify opportunities for improvement.
• Build automated model retraining pipelines.
• Perform feature engineering and analyze historical data.
• Investigate model errors and develop solutions for underperforming scenarios.
• Determine when additional features are appropriate versus when a specialized model may be needed.
• Develop predictive models using structured and historical data.
• Build data pipelines supporting model training and feature engineering.
• Track experiments and model performance.
• Deploy and integrate ML applications into production environments.

Preferred Background

• Machine Learning Engineering experience.
• Experience with predictive modeling.
• Strong Python experience.
• Experience with XGBoost, LightGBM, or similar ML algorithms.
• Feature engineering and model evaluation.
• Experience building ML/data pipelines.
• Production ML / MLOps experience.
• AWS experience.
• MLflow or similar experiment-tracking experience.
• Strong problem-solving skills and the ability to investigate why models succeed or fail.

Environment

AWS | Python | XGBoost | MLflow | DVC | FastAPI | HashiCorp Nomad

This is a great opportunity for someone who wants to work on real predictive AI problems with measurable business impact, rather than purely theoretical ML research.

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