ML Engineer - Computer Vision
Role summary
We are seeking a full-time, in-person Machine Learning Engineer specializing in Computer Vision to join our early-stage startup focused on fighting climate change. You will leverage autonomous vehicle technology, sensors, and perception AI to manage forests with high accuracy. Responsibilities include developing data pipelines, training neural networks on satellite imagery, visualizing data, and integrating geospatial data into our platform. This role requires 3+ years of experience with ML algorithms and Python, deep learning model training in PyTorch or Jax, and understanding of the model life cycle. Cloud environment experience and strong statistical/geometric knowledge for computer vision are also essential.
About Us
We are currently building our AI team and hiring for a full-time, in-person Machine Learning Engineer - Computer Vision with the ambition to help get things done. This is a great opportunity for someone who wants to apply their skills towards fighting climate change, wants to experience the high-pace, high-ownership setting of an early startup, and can start onboarding immediately or in the next few weeks.
Climate change is one of the most meaningful challenges of our generation, and every possible solution put forward by the United Nations IPCC requires scaled carbon sequestration. Forests are critical for this reason, but they need to be managed for a growing set of values including sustainable wood production, carbon, biodiversity, and wildfire risk mitigation. We are an MIT climatetech startup applying autonomous vehicle technology to solve this problem, utilizing cutting edge sensors and perception AI to build the modern tooling needed to manage forests with great accuracy and confidence.
Job brief
We are looking for a machine learning engineer who is motivated in building high-impact technology that will make the world a better place. Responsibilities may range from implementing our vision data pipeline, training neural networks on satellite imagery, visualizing data in order to make decisions on how to improve the system, and incorporating large geospatial data and outputs in our platform. You will work with the whole team and help everywhere you are willing and able. This is an opportunity to join a well-positioned, ambitious technology startup making a large impact fighting climate change.
Responsibilities (may include)
Research and implement algorithms from machine learning and satellite imagery modeling papers, and make effective use of state-of-the-art pretrained models
Design, implement, and integrate data pipelines, deployed to AWS using Docker
Identify bottlenecks and improve software efficiency
Requirements
3+ years proven experience with machine learning algorithms and Python.
Has trained and applied deep learning models in Pytorch or Jax for some part of the perception stack such as classification, detection, segmentation, tracking, or mapping.
Understands most parts of the model life cycle from training to deployment.
Uses DevOps and Python best practices.
Core statistical and geometric knowledge for computer vision and machine learning.
Can work in cloud-like environments applying system administration and system architecture skills.
Nice to haves
Has applied computer vision in a related field such as remote sensing, geospatial data processing, robotics, self driving, medical, or defense.
Has worked with data modalities such as video, multispectral image stacks, or lidar.
Has experience with data pipelines for ingesting, cleaning, transforming, preprocessing, and postprocessing data.
Can apply full stack, backend, and frontend knowledge to integrate machine learning models into web apps.
Has used common MLOps tooling such as Weights and Biases, Pytorch Lightning, MLFlow, or Python Ray.
Can apply advanced frequentist or bayesian statistical modeling to a scientific field.
Is familiar with modern computer vision models such as VLLMs, Transformers, Vision Transformers, Gaussian Splats, or Neural Rendering
