Computer Vision and AI Software Engineer - Defense & Robotics

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.

Universal Labeler – The Data Engine of Physical AI

We are building the Universal Labeler — a system that generates data at scale, not just labels it.

Instead of labeling frames, we label reality across: 2D ↔ 3D ↔ multi-sensor domains

This enables:

• Massive scaling of labeled data
• Cross-sensor ground truth transfer (RGB → Thermal → Radar)
• Continuous generation of training data from real environments

The goal:

a machine for building foundation models across all sensors and conditions.

Algolight Edge – Real-Time AI in the Field

We build Algolight Edge — systems that take AI out of the lab and into the real world:

• Real-time pipelines running on edge devices (Jetson-class and beyond)
• Full-stack optimization (latency, memory, throughput)
• Multi-sensor processing under real-world constraints (noise, motion, instability)

This is where models meet reality — and have to actually work.

This is not a tool.

This is infrastructure for the next generation of physical AI.

R&D Team Mission

You will join a small, high-intensity R&D team working across:

• Universal Labeler systems (data generation, labeling pipelines, geometry-based truth)
• Edge AI systems (real-time deployment on Jetson and similar platforms)
• Video enhancement systems (real-world, degraded data, real-time + forensic pipelines)

This is not a narrow role.

You will operate across the full stack:

data → models → systems → deployment

🚀 What You’ll Be Responsible For

Full-Stack AI Problem Solving

● Take open-ended problems and drive them to working solutions.

● Move between domains: computer vision, backend systems, pipelines, infrastructure.

● Build systems, not just models.

Universal Tagger Development

● Implement components in the 2D ↔ 3D ↔ 2D pipeline.

● Work on geometry-based labeling, projection, and dataset generation.

● Build automation flows for large-scale dataset creation.

● Integrate multiple models into unified labeling outputs.

Edge AI Systems

● Build real-time AI pipelines running on edge devices (Jetson-class hardware).

● Optimize inference pipelines (latency, memory, throughput).

● Work under real-world constraints (noise, instability, limited compute).

Video Enhancement Systems

● Develop and improve video enhancement pipelines.

● Handle degraded, noisy, and real-world sensor data.

● Combine classical algorithms with ML-based approaches.

Research to Production

● Take ideas from concept → working prototype → deployed system.

● Iterate based on real-world data, not just offline benchmarks.

● Work closely with field and operational constraints.

🎯 What We Are Looking For

Core Mindset (Critical)

We are not primarily looking for years of experience.

We are looking for someone who:

● Uses AI as a primary working tool

● Can enter any domain (CUDA / CV / backend / infra / etc.) and figure things out

● Learns extremely fast

● Most importantly — delivers results

Required Capabilities

● Hands-on programming experience (Python required, C++ / CUDA is a strong plus)

● Experience building real systems (not only academic work)

● Ability to take ambiguous problems and produce working solutions

● Strong debugging and problem-solving skills

● Comfort working across multiple technical domains

Technical Strengths

● Computer vision / image processing (classical and/or deep learning)

● Detection, Segmentation, Tracking

● Systems thinking (pipelines, data flow, integration)

● Basic understanding of ML pipelines and model usage

● Familiarity with Linux environments and development workflows

⭐ Strong Advantages

● Experience with CUDA / GPU programming

● Experience with edge devices (Jetson, embedded systems)

● Background in signal processing or physics-based systems

● Experience working with real-world sensor data (not just clean datasets)

● Exposure to multi-sensor systems

🌟 What Awaits You at Algolight

● Direct work with the CTO and core research leadership

● Ownership over real systems — not isolated tasks

● Exposure to the full stack:

AI × systems × sensors × deployment

● Extremely fast learning curve

● Work on problems that exist in the real world — not just benchmarks

● A team that values:

execution, depth, and getting things to actually work

🔎 Sensors & Domains You’ll Work With

● Visible (RGB), Thermal (IR), Low-light

● Radar and additional sensing modalities

● Multi-sensor fusion pipelines

● Real-time edge systems