Applied AI Researcher - Satellite Team

At Algolight, we live and breathe the future of artificial intelligence and the physical world.

Our mission it two-folder:

On the civilian side, we build labeled 3D information layers from all types of sensors—for smart cities, drones, autonomous vehicles, infrastructure, public safety, and far beyond.

On the defense side, we bring true real-time intelligence to the edge—anywhere, for any sensor, at any point on the map—enabling smart, real-time decisions in the field.

VISINT & Multi-Sensor Intelligence Platform

At Algolight, we are building large-scale AI systems that transform satellite and aerial data (EO, SAR, and beyond) into actionable intelligence.

These systems span:

• Massive data ingestion pipelines from spaceborne and airborne sensors
• Advanced AI/ML pipelines for detection, change analysis, and scene understanding
• Distributed compute environments for training, inference, and analytics
• Operational systems deployed in real-world, mission-critical environments

As an Applied AI Researcher, you will work on some of the hardest problems in real-world AI—where data is noisy, labels are scarce, and performance directly impacts operational systems.

🚀 What You’ll Be Responsible For

Applied Research in Computer Vision & GeoSpatial AI

• computer vision
• geospatial AI
• multi-sensor learning
• Develop algorithms and models for:
• object detection
• segmentation
• change detection
• anomaly detection

Model Development & Evaluation

• Design, implement, and evaluate machine learning and deep learning models
• Work with architectures such as:
• CNNs
• Transformers
• Vision-Language Models
• Continuously improve performance through:
• experimentation
• error analysis
• iterative refinement

Multi-Temporal & Time-Series Analysis

• Develop methods for analyzing time-varying data, including:
• change detection across temporal sequences
• trend and anomaly detection
• Handle real-world challenges such as:
• seasonal variation
• sensor drift
• environmental changes

Multi-Modal Fusion

• Build models that combine multiple data sources:
• EO (optical imagery)
• SAR (radar imagery)
• metadata and contextual signals
• Develop fusion strategies:
• early / mid / late fusion
• representation learning across modalities

Working with Real-World Data Challenges

• Handle complex datasets characterized by:
• noise (e.g., SAR speckle)
• domain variability
• incomplete or sparse labels
• Develop robust solutions under:
• data scarcity
• domain shift
• imperfect annotations

Experimentation & Research Rigor

• Design and run experiments:
• define hypotheses
• evaluate performance
• analyze results
• Read, understand, and apply state-of-the-art research
• Translate research ideas into working solutions

From Research to Production

• Work closely with:
• MLOps engineers
• data engineers
• system teams
• Ensure smooth transition of models from:
• research → validation → deployment

User & Operational Collaboration

• Collaborate with:
• analysts
• interpreters
• domain experts
• Validate model outputs against real-world use cases
• Incorporate operational feedback into model design

🎯 What We Are Looking For

Required Experience

• 3+ years of experience in Machine Learning / Deep Learning
• Proven experience developing models for real-world problems
• Strong background in Computer Vision and/or Image Processing

Core Technical Skills

• Programming:
• Python (mandatory)
• ML Frameworks:
• PyTorch / TensorFlow
• Model Development:
• Experience training, evaluating, and debugging models
• Research Skills:
• Ability to read, understand, and implement academic papers

Core Strengths

• Strong analytical and problem-solving skills
• Ability to work with complex, messy datasets
• Experimental mindset and attention to detail
• Ability to operate in a multi-disciplinary environment

⭐ Strong Advantages

• Experience with satellite data (EO / SAR)
• Understanding of SAR processing:
• speckle noise, calibration, dB domain
• Experience with:
• change detection
• time-series / multi-temporal analysis
• Experience with:
• Transformers
• Vision-Language Models
• Experience with:
• multi-modal learning
• data fusion
• Familiarity with:
• MLOps tools (MLflow / W&B)
• Experience in operational or real-world systems

🌟 What Awaits You at Algolight

• Work on real-world intelligence problems with direct operational impact
• Access to rare multi-sensor datasets (including satellite data)
• Deep involvement in:
• AI × data × physics × real-world systems
• Collaboration with top-tier researchers, engineers, and domain experts
• A culture that values:
• depth
• rigor
• real impact over benchmark performance

🔎 Systems & Data You’ll Work With

• Satellite imagery (EO, SAR)
• Multi-sensor datasets (optical, radar, metadata)
• Time-series and multi-temporal data
• Large-scale geospatial data pipelines