Satlantis is a leading-edge company specializing in high-performance satellite technology and data processing. We are at the forefront of innovation, developing advanced solutions for Earth observation and space exploration. Join our dynamic team in Gainesville, Florida, and contribute to groundbreaking projects that shape the future of satellite technology. For more information about the company, please visit www.satlantis.com .
Position SummaryWe are seeking a highly motivated and experienced
Senior Computer Vision Engineer with strong technical leadership to drive Satlantis US's computer vision and imagery understanding initiatives. The ideal candidate combines deep expertise in modern vision systems with pragmatic delivery: you will design, develop, evaluate, and deploy computer vision models and pipelines that operate on large-scale satellite imagery and geospatial data products.
This is a hands-on role where you will own vision workstreams end-to-end-from problem definition and dataset strategy to model development, production deployment, performance optimization, and iteration-while setting engineering standards, mentoring teammates, and partnering closely with engineering, product, and mission teams. You will help ensure our computer vision systems are accurate, robust, scalable, and operationally effective in real-world Earth-observation workflows.
What you'll do: - Own your developments. Lead high-impact computer vision initiatives such as segmentation, object detection, classification, image matching, semantic retrieval, change detection, tracking, and anomaly detection over satellite imagery and derived geospatial products, delivering measurable improvements in model quality and operational outcomes.
- Translate problems into vision systems. Convert customer needs, mission requirements, and research goals into well-scoped computer vision problems, define success metrics and KPIs (e.g. precision/recall, mAP, IoU, F1, latency, throughput, memory footprint), and establish acceptance criteria and validation plans.
- Design datasets that win. Drive dataset strategy for vision applications, including annotation protocols, tiling and sampling strategies, class balance, hard-negative mining, augmentation policies, domain-shift analysis, and label-quality audits. Establish repeatable dataset versioning and documentation practices.
- Build robust training and evaluation pipelines. Implement reproducible experimentation, benchmarking, ablation studies, and error-analysis workflows for computer vision models, including geospatially aware evaluation where applicable.
- Advance model architectures. Develop and improve state-of-the-art computer vision approaches, including CNNs, transformers, encoder-decoder architectures, self-supervised learning, multi-modal fusion, and foundation-model adaptation for remote sensing imagery. Optimize solutions for real operational constraints such as image resolution, viewing conditions, atmospheric noise, and multi-temporal data.
- Operationalize vision models. Partner with software and platform engineers to productionize vision systems, including model packaging, inference optimization, deployment pipelines, monitoring, drift detection, versioning, rollback strategies, and performance tuning across heterogeneous compute environments.
- Raise the engineering bar. Set standards for code quality, reproducibility, model validation, benchmarking, documentation, and peer review. Write clear technical design documents and decision memos that align stakeholders and accelerate execution.
Skills and experience (required): - Bachelor's, Master's, or PhD in Computer Science, Computer Engineering, Electrical Engineering, Remote Sensing, Robotics, or a related field.
- 3+ years of professional experience in computer vision, machine learning, or applied AI, including delivering vision models into production or operational workflows.
- Strong proficiency in Python for machine learning and computer vision workflows; ability to write clean, maintainable, and well-tested code.
- Deep knowledge of computer vision fundamentals, including image representations, feature extraction, geometric reasoning, dense prediction, detection, segmentation, and model evaluation.
- Hands-on experience with deep learning frameworks such as PyTorch (preferred) or TensorFlow, and practical experience implementing modern vision architectures.
- Strong understanding of training and inference optimization, including data loading efficiency, batching, mixed precision, model compression, and performance-aware experimentation.
- Experience working with large-scale imagery or visual datasets and building pipelines that are reliable and reproducible.
- Strong communication skills, with the ability to explain complex technical trade-offs clearly to cross-functional stakeholders.
Nice to have (preferred) - Geospatial / satellite domain experience: GDAL, Rasterio, projections/CRS, tiling strategies, GeoTIFF/COG/NetCDF, STAC/PgSTAC, and geospatial image-quality considerations.
- Remote-sensing computer vision: experience with multi-spectral or panchromatic imagery, super-resolution, image fusion, orthorectification-aware workflows, and change detection in Earth-observation contexts.
- Spatiotemporal vision modeling: time-series imagery, temporal fusion, motion/change analysis, event detection, or tracking across repeated satellite captures.
- MLOps / production AI: model serving, monitoring, experiment tracking (e.g. W&B, MLflow, CometML), orchestration (Airflow, Argo, ZenML), and lifecycle management.
- Cloud & compute: experience training and running inference on AWS/GCP/Azure and on-prem HPC/cluster environments, including SLURM-managed GPU/CPU fleets and Kubernetes-based infrastructure; strong understanding of containers, distributed training, GPU scheduling, storage/performance bottlenecks, and cost/performance tuning.
- Foundation models for vision / Earth observation: fine-tuning, embedding extraction, retrieval systems, transfer learning, promptable models, and multimodal representation learning.
- C++ or performance-oriented deployment experience: OpenCV, ONNX, TensorRT, Triton Inference Server, CUDA optimization, or edge/real-time inference workflows.
- Familiarity with data governance and quality frameworks, including lineage, validation checks, and dataset documentation.
Work Authorization:This role will
not sponsor any employment visas. Candidates must have and maintain unrestricted legal authorization to work in the U.S. now and in the future, without requiring employer-sponsored visa support.
Location & Work Model:Full-time, in-person position in Gainesville, Florida. You'll work closely with engineering and business teams on impactful, real-world satellite analytics and AI systems-helping deliver reliable, scalable capabilities that push forward the state of the art in Earth observation.
Why Join Satlantis? - Be part of a pioneering company at the forefront of space technology.
- Work on challenging and impactful projects that have real-world applications.
- Collaborate with a team of brilliant and passionate engineers and scientists.
- Competitive salary and benefits package.
- Opportunity for professional growth and development in a rapidly expanding industry.
- Enjoy the vibrant community and quality of life in Gainesville, Florida. Learn more at https://www.visitgainesville.com/ .
