Machine Learning Engineer

CETO Innovation is developing predictive maintenance technology for district heating infrastructure.

Our solution combines an in-pipe inspection probe, sensing technology, utility data, and predictive analytics to help operators understand underground pipeline condition before failures happen.

We are looking for a Machine Learning Engineer to develop CETO's predictive model for utility pipes.

Turn pipe-expert knowledge, utility data, and inspection signals into quantified risk models that support real maintenance decisions.

What you'll work on Develop CETO’s predictive maintenance model for district heating pipes

Translate pipe-expert knowledge into quantified model features, rules, weights, and risk scores

Work with utility data, inspection data, operational logs, pipe metadata, and environmental factors

Build data pipelines, feature extraction, data-quality checks, and validation workflows

Model degradation drivers such as corrosion, welding defects, biofouling, thermal stress, and water chemistry

Combine domain rules, statistical models, and ML methods into a practical decision-support framework

Create calibrated probabilistic risk estimates, not only point predictions

Evaluate false positives, false negatives, uncertainty, and ranking consistency

Integrate field-test results from CETO’s probe to recalibrate and improve the model

Collaborate with hardware and software engineers so sensor data becomes reliable model input

Help turn model outputs into useful maintenance recommendations for utility teams

What we're looking for Strong experience with Python and data science/ML workflows

Experience building predictive models from messy, real-world data

Good understanding of statistics, model validation, uncertainty, and performance metrics

Experience with time series, anomaly detection, classification, regression, or risk scoring

Ability to turn domain expertise into structured features, assumptions, and model logic

Experience with data pipelines, cleaning, feature engineering, and experiment tracking

Comfort working with limited, imperfect, or partially labelled datasets

Experience with Git and GitHub

Experience with infrastructure, energy, utilities, industrial systems, sensor data, physics-informed ML, or predictive maintenance is a plus

How we work You will join a small, hands-on startup team where everyone takes ownership and works close to the problem. In this role, you will collaborate with pipe experts, hardware engineers, and software engineers to transform expert judgement and field data into a model that utilities can trust.

#J-18808-Ljbffr