Machine Learning Engineer — Computer Vision

Innomium; Northern (KY); Hybrid; USD 150,000 - 210,000; Full time

Benefits offered by this job

Remote-first flexibility

Health and wellness stipends

Annual learning budget

Equipment and setup stipend

Retirement contributions

Job summary

Innomium is seeking a Machine Learning Engineer in computer vision to own vision work from data and baseline design through training, error analysis, model export, runtime profiling, and pilot evidence. Projects may involve object detection, segmentation, event logic, edge deployment, or adaptation of public releases such as Sentinel, Vantage, and Ember.

This role emphasizes evaluating scenes, understanding occlusion, and determining when post-processing improves outcomes, partnering with

Qualifications

• Professional experience developing and evaluating modern computer-vision systems.
• Strong Python, PyTorch, data-pipeline, and experiment-management skills.
• Practical understanding of detection metrics, dataset bias, augmentation, and error analysis.
• Experience taking models into a runtime outside the training environment.
• Ability to reason about latency, memory, hardware, privacy, and operational consequences.
• Evidence of reproducible technical work through code, models, papers, demos, or shipped systems.
• Strong written communication and careful experiment documentation.
• Ability to work effectively in a remote environment with autonomy and accountability.

Responsibilities

• Design representative datasets, splits, annotation guidance, and scene-level evaluation protocols.
• Train, adapt, distill, and compare detection or segmentation models for target environments.
• Perform structured error analysis across camera, condition, class, scale, and failure mode.
• Export and validate ONNX or other deployment artifacts and profile complete inference pipelines.
• Collaborate on temporal logic, event semantics, review interfaces, and monitoring.
• Document model lineage, data limitations, runtime assumptions, and production acceptance evidence.
• Respect privacy, safety, and the limits of computer vision in high-consequence workflows.
• Communicate trade-offs clearly to engineering and non-engineering collaborators.

Skills

Python

PyTorch

Data pipelines

Experiment management

Latency analysis

Documentation

Remote collaboration

Job description

Develop and deploy scene-specific vision systems across data, training, evaluation, packaging, edge inference, and operational alert behavior.

Innomium is an applied AI research and engineering company that turns ambitious technical ideas into dependable, production-ready systems.

We bring together AI research, product engineering, data, cloud infrastructure, evaluation, and operational delivery within one accountable program. Our teams work with startups, product companies, and enterprises to build custom AI models, software products, deployment pipelines, integrations, and reproducible evaluation systems.

Our work spans language models, AI agents, computer vision, retrieval systems, cloud and edge deployments, open research releases, and engineering contributions. Through Innomium Arena, we also create structured opportunities for builders to contribute to challenging technical projects. Through Innomium Compute, we provide on-demand GPU capacity for training and inference.

We focus on measurable outcomes, inspectable evidence, and software that teams can operate and improve—not prototypes that stop at the demonstration stage.

The Role

As a Machine Learning Engineer in computer vision, you will own vision work from data and baseline design through training, error analysis, model export, runtime profiling, and pilot evidence.

The Innomium Vision program publishes compact detection artifacts and applies the same discipline to real operating environments. We treat camera conditions, data quality, runtime constraints, and workflow behavior as part of the model problem. Projects may involve object detection, segmentation, event logic, edge deployment, or adaptation of public releases such as Sentinel, Vantage, and Ember.

This role is not only about improving a headline score. You will investigate which scenes fail, how object scale and occlusion change outcomes, whether post-processing helps, and what an alert should mean to the operator.

You will partner with research, data, product, and infrastructure engineers.

What Strong Performance Looks Like

You create evaluation protocols that expose the difficult tail, produce reproducible training and inference artifacts, and explain the accuracy–latency–size trade-off clearly.

You can move a promising model into a bounded pilot without overstating what the evidence proves. Your experiment records, model cards, and runtime assumptions are clear enough for another engineer to reproduce and challenge.

Over time, you improve Innomium’s vision delivery system: dataset quality, error analysis habits, export validation, and operational monitoring for production camera workflows.

How We Work

Innomium operates through small, accountable teams with direct access to the technical problem.

We value:

• Clear ownership and reliable execution.
• Written decisions and reviewable technical reasoning.
• Measurable acceptance criteria.
• Honest communication about risks and limitations.
• Practical solutions over unnecessary complexity.
• Documentation and handover from the beginning of a project.
• Engineering decisions connected to user and operating outcomes.

Remote collaboration requires dependable communication, thoughtful handoffs, and agreed working-hour overlap with the relevant delivery team.

Compensation and Benefits

Compensation range: $150,000–$210,000 USD (base), depending on experience, location, and engagement type. Total compensation may include performance-based bonuses or equity participation where applicable.

Employment arrangement: Full-time

Location and working hours: Remote. United States preferred; international candidates are considered subject to work authorization, contracting or employment availability, and required overlap with team working hours.

Health and wellness: Medical, dental, and vision coverage (or equivalent stipend for international contractors), plus access to mental health and wellness support programs.

Paid time off: Flexible paid time off policy, including vacation, sick leave, and company holidays. Parental leave provided in accordance with local regulations and role type.

Professional development: Annual learning and development budget for courses, certifications, books, and conferences. Support for attending relevant industry events and technical communities.

Equipment and remote-work support: Company-provided laptop and necessary development equipment. Monthly stipend for internet and home-office setup where applicable. Access to required software and cloud tools.

Additional benefits: Retirement or pension contributions where applicable, remote-first flexibility, and potential performance-based bonuses or equity participation depending on role and engagement type.

What You Will Own

The work this role is expected to own.

• Design representative datasets, splits, annotation guidance, and scene-level evaluation protocols.
• Train, adapt, distill, and compare detection or segmentation models for target environments.
• Perform structured error analysis across camera, condition, class, scale, and failure mode.
• Export and validate ONNX or other deployment artifacts and profile complete inference pipelines.
• Collaborate on temporal logic, event semantics, review interfaces, and monitoring.
• Document model lineage, data limitations, runtime assumptions, and production acceptance evidence.
• Respect privacy, safety, and the limits of computer vision in high-consequence workflows.
• Communicate trade-offs clearly to engineering and non-engineering collaborators.

Required Qualifications

Capabilities and experience that support success in this role.

• Professional experience developing and evaluating modern computer-vision systems.
• Strong Python, PyTorch, data-pipeline, and experiment-management skills.
• Practical understanding of detection metrics, dataset bias, augmentation, and error analysis.
• Experience taking models into a runtime outside the training environment.
• Ability to reason about latency, memory, hardware, privacy, and operational consequences.
• Evidence of reproducible technical work through code, models, papers, demos, or shipped systems.
• Strong written communication and careful experiment documentation.
• Ability to work effectively in a remote environment with autonomy and accountability.

Preferred Qualifications

Valuable adjacent experience, but not a substitute for the core requirements.

• Experience with YOLO-family models, ONNX Runtime, TensorRT, OpenVINO, or edge accelerators.
• Background in industrial, logistics, safety, or embedded-vision environments.

Follow Innomium for hiring and company updates.

We’re actively growing across engineering, research, and product. Follow us on LinkedIn, GitHub, and Hugging Face to hear about new roles, releases, and the work we ship in public—without waiting for a careers-page refresh.