Echo Labs - Director of Modelling

Introduction

Echo Labs is building a scientific, and technical foundation for ecological intelligence: a multimodal system to measure, model, and forecast Ecosystem Condition as a dynamic property. We are a collaborative and interdisciplinary team of scientists and engineers engaged in a planetary moonshot – with a public good mission, operating like a start up.

We are a new Focused Research Organization (FRO) supported by Convergent Research and funded by the Advanced Research and Invention Agency to pursue high-risk, high-reward science in the public interest.

Echo Labs is seeking a Director of Modelling to design and build the data pipelines and modelling infrastructure that translate raw ecological signals into testable models of ecosystem condition. You will help turn ecological hypotheses into experiments, experiments into validated models, and validated models into decision‑grade outputs. As we explore what is possible with new measures and representations of ecosystem state, you will be responsible for evaluating progress.

You will help define success in a system that is complex to model using results that are challenging to interpret. This role reports to the Chief Technology Officer and works in close partnership with the Chief Science Officer. This is a rare opportunity to architect the conceptual and technical frameworks that define how ecosystems are measured and modelled.

• Experiment design & strategy:
  • Design experiments that isolate signal from noise in ecological data, and design tests that can increment progress forward when results are ambiguous.
  • Work with the CTO to prioritize intermediate outputs, publication opportunities, and risks.
  • Work iteratively with the CSO to test and refine modelling approaches and uncertainty quantification that can be interpreted in ecological context.
  • Define a standardization strategy that ensures model comparability across ecosystems and geographies.
• Research execution:
  • Translate ecological hypotheses from the CSO into testable ML experiments.
  • Implement, train, and evaluate models that characterise ecosystem condition across multiple measurement modalities. This will range from implementing classic machine learning to fine‑tuning modern deep learning models.
  • Contribute to research outputs: white papers, technical documentation, and peer‑reviewed publications.
• Maintain infrastructure:
  • Working with a growing team under your management, you will develop and maintain the technical stack to ensure reproducible, rapid experimentation.
• Team and growth:
  • Work with the leadership team to hire and lead additional engineers and data scientists.
  • Evaluate and communicate needs and opportunities to creatively accelerate progress.
  • Establish technical standards, code review practices, and documentation norms.
• Profile (You Have)
  • 8+ years in machine learning, data engineering, or applied research with end‑to‑end system ownership.
  • Track record with machine learning infrastructure and experiment tracking, model versioning, reproducible pipelines.
  • Experience with multimodal data. At minimum two of time‑series sensor data, audio/acoustic data, imagery, or geospatial/remote sensing.
  • Experience with cloud platforms (AWS or GCP) and MLOps tooling.
  • Demonstrated ability to scope problems, make architectural decisions, and deliver without close supervision.
  • Strong scientific communication. You have the ability to explain technical choices across domains and document work for reproducibility.
• Highly Valued
  
  Experience:
  
  
  • Background in ecology, environmental science, Earth observation, or prior work with ecological datasets.
  • Prior work with bioacoustic data, eDNA, or biodiversity monitoring systems.
  • Experience hiring and managing technical teams at early‑stage organizations.
  • Contributions to open‑source ML or scientific software projects.

£145,000 - £168,000 a year

Title commensurate with experience:
Senior Director / Chief Technology Officer possible.

Bonus:
Performance‑based

In the first six months, you'll help deliver an architecture for Echo's modeling platform. Core data pipelines will be operational using previously‑generated datasets. First experiments…