Principal Autonomy & Perception Systems Engineer

Job Description

Title: Principal Autonomy & Perception Systems Engineer

Company Name: Hexon

Vacancy: --

Age: At least 18 years

Job Location: Anywhere in Bangladesh

Salary: Tk. 100000 - 160000 (Monthly)

Experience:

Published: 2026-08-15

Application Deadline: 2026-08-31

Education:

Requirements:

Skills Required:

Additional Requirements:

• Age At least 18 years

About Hexon

Hexon Technology is building ATLAS, an edge-based defense intelligence and autonomy platform designed to transform multiple sensor feeds into a persistent understanding of the physical environment.

ATLAS combines computer vision, multi-object tracking, sensor fusion, persistent identity, geospatial visualization, event timelines, threat assessment, and decision support into a unified operational platform.

Our long-term objective is to connect this intelligence layer with multiple sensors, autonomous systems, and defensive effectors.

We are looking for the technical leader who can help take ATLAS from its current working system into a production-grade defense platform.

The Role

We are seeking an exceptional Principal Autonomy & Perception Systems Engineer to become the technical lead for ATLAS.

This is not primarily a management position.

You will be expected to understand the existing codebase, write production code, make architectural decisions, solve difficult perception and tracking problems, establish engineering standards, and lead a small multidisciplinary engineering team.

You will work directly with the founder.

What You Will Own

You will have technical ownership across the ATLAS pipeline:

Sensors → Detection → Tracking → Sensor Fusion → Persistent Identity → World Model → Threat Assessment → Decision Support → Effector Integration

Responsibilities include:

• Assess and take ownership of the existing ATLAS architecture and codebase

• Design the next-generation ATLAS system architecture

• Build reliable real-time multi-object tracking

• Improve multi-camera tracking and cross-camera identity persistence

• Develop sensor-fusion architecture for EO/IR, radar, RF and other sensors

• Build and maintain a persistent world model

• Develop geospatial and temporal representations of tracked entities

• Architect edge-first processing for disconnected and bandwidth-constrained environments

• Optimize perception workloads for GPU/edge hardware

• Define APIs and interfaces for sensors, autonomous systems, and external effectors

• Establish testing, simulation, logging, observability and system-health standards

• Help architect the future ATLAS decision layer

• Recruit and technically lead additional engineers as the team grows

• Take ATLAS from prototype through field testing, hardening and production

Required Experience

We care significantly more about demonstrated capability than credentials.

Strong candidates should have deep experience in several of the following:

• Robotics or autonomous systems

• Computer vision

• Multi-object tracking

• Multi-camera tracking

• Re-identification

• Sensor fusion

• State estimation

• Real-time perception systems

• C++ and Python

• OpenCV

• PyTorch

• NVIDIA CUDA/TensorRT

• Linux

• Docker

• Distributed or event-driven systems

• Edge computing

• Geospatial systems

• Real-time data pipelines

You should understand the difference between a detection, observation, track, identity and persistent real-world entity and be capable of designing the architecture connecting them.

Highly Desired

Experience with:

• Defense or aerospace systems

• Counter-UAS

• ISR

• Radar/EO/IR integration

• ROS2

• Kalman filtering and probabilistic tracking

• Multi-target/multi-sensor tracking

• NVIDIA Jetson

• MAVLink / PX4 / autonomous platforms

• GIS and coordinate transformations

• Simulation and synthetic sensor environments

• Human-in-the-loop autonomy

• Command-and-control systems

• Degraded or disconnected operational environments

The Person We’re Looking For

We’re specifically looking for a builder.

You should be comfortable opening an unfamiliar repository, running the system, identifying architectural weaknesses, debugging real-time perception problems, writing the critical code yourself, and then directing other engineers toward a coherent architecture.

You should be able to move comfortably between:

Algorithm → Code → Sensor → Edge Computer → Network → Operator Interface → Physical World

We do not need someone who simply manages Jira tickets or coordinates a development team.

We need someone capable of becoming the technical authority for ATLAS.

Initial Mission

Your first responsibility will be to evaluate the existing ATLAS platform and determine:

1. What should be retained

2. What should be refactored

3. What needs to be rebuilt

4. What prevents reliable persistent tracking and identity today

5. What architecture is required for additional sensor modalities

6. What is required to move from the current system to a fieldable product

7. What engineering team should be built around the platform

From there, you will lead execution of that roadmap.

Why Hexon

This is an opportunity to take technical ownership of an ambitious defense technology platform at an early stage.

You won’t be joining a large engineering organization to maintain one component.

You will help determine what ATLAS becomes and build the engineering organization responsible for getting it there.

If you’ve built real-world perception, robotics, autonomy, tracking, or sensor-fusion systems and want substantially more technical ownership, we want to hear from you.