Machine Learning Engineer

Overview

Lightningware is looking for a Machine Learning Engineer focused on AI Safety and Moderation to help build the next generation of intelligent safety systems for our live video platform, Thundr and future products.

Thundr operates a real-time video chat service where trust, safety, and moderation are critical to the user experience. We already have moderation systems in production, and we are looking for an engineer who can help us significantly expand their capabilities. Improving detection accuracy, reducing response times, scaling our infrastructure, and researching new ways AI can make real-time online interactions safer.

This role will work across computer vision, multimodal AI, machine learning infrastructure, real-time inference, and applied AI research. You will help develop systems capable of analyzing live video, images, audio, text, and behavioral signals to identify policy-violating or unsafe activity at scale.

You will also play an important role in an upcoming Lightningware project where we are building new moderation technology from the ground up.

This is a highly applied engineering role. We are looking for someone who enjoys taking emerging AI capabilities and turning them into reliable production systems operating at significant scale.

Responsibilities

• Design and build machine learning systems for real-time moderation of live video interactions.
• Develop computer vision and multimodal models capable of detecting unsafe, inappropriate, abusive, or policy-violating content.
• Research and evaluate emerging AI models and techniques that could improve Thundr’s moderation and safety capabilities.
• Improve existing moderation systems by optimizing precision, recall, latency, throughput, and cost.
• Develop systems for model evaluation, benchmarking, dataset creation, labeling, and continuous improvement.
• Build scalable inference infrastructure capable of processing large volumes of real-time user activity.
• Design moderation architectures that intelligently combine automated detection, confidence thresholds, escalation systems, and human review.
• Investigate techniques for detecting adversarial behavior and users attempting to circumvent automated moderation systems.
• Deploy and monitor machine learning models in production environments.
• Develop tools that allow our team to understand model decisions, review incidents, and improve moderation policies.
• Collaborate closely with engineering and leadership teams on new products and safety initiatives.
• Maintain strong documentation around model architecture, evaluation methodology, datasets, deployment systems, and moderation experiments.

Qualifications

• Strong experience building and deploying machine learning or AI systems in production.
• Bachelor’s degree in Computer Science, Machine Learning, Artificial Intelligence, Engineering, or a related technical field, or equivalent practical experience. Master’s degree is a plus, but not required.
• 3+ years of professional experience in machine learning engineering, applied AI, computer vision, or a related field, with experience deploying ML systems into production.
• Strong Python programming skills and experience with modern ML frameworks such as PyTorch, TensorFlow, or similar technologies.
• Experience with computer vision, multimodal AI, NLP, classification systems, or content moderation.
• Understanding of modern deep-learning architectures including transformers and vision models.
• Experience designing model evaluation frameworks and working with metrics such as precision, recall, F1, false-positive rates, and false-negative rates.
• Experience building data pipelines for training, evaluation, and inference.
• Familiarity with cloud infrastructure such as AWS, GCP, or Azure.
• Experience deploying scalable model inference systems through APIs, containers, GPUs, or distributed infrastructure.
• Strong understanding of software engineering principles and the ability to build reliable production systems.
• Ability to independently research new AI techniques, evaluate their usefulness, and turn promising approaches into working prototypes.

Preferred qualifications

• Experience building trust & safety, content moderation, fraud detection, abuse prevention, or platform integrity systems.
• Experience processing live or near-real-time video streams.
• Experience with technologies such as OpenCV, FFmpeg, WebRTC, or video-processing pipelines.
• Experience working with multimodal foundation models or vision-language models.
• Experience with GPU optimization and high-throughput inference.
• Experience designing annotation or labeling pipelines for sensitive datasets.
• Familiarity with model serving technologies such as Triton, ONNX Runtime, TensorRT, Ray, or similar systems.
• Experience with MLOps, model monitoring, experiment tracking, and automated retraining pipelines.
• Experience researching emerging AI techniques and rapidly prototyping new approaches.

What Success Looks Like

In this role, you will help Thundr create moderation systems that can operate intelligently and reliably across millions of real-time interactions.

Success means:

• Detecting harmful content more accurately.
• Reducing false positives that negatively affect legitimate users.
• Identifying unsafe behavior faster.
• Building moderation systems that can scale as Thundr grows.
• Giving our safety teams better tools and signals for making decisions.
• Continuously evaluating emerging AI technology and incorporating meaningful improvements into our platform.
• Establishing reusable AI safety infrastructure that can support both Thundr's existing live service and future products.

Why Lightningware

At Lightningware, you’ll work on real-world AI challenges where your work directly impacts the safety and experience of our users.

You’ll help improve the moderation systems powering Thundr’s live video platform, while researching and implementing new approaches across computer vision, multimodal AI, and real-time moderation.

We also have an upcoming project in development where you’ll have the opportunity to help build its moderation and ML systems from the ground up.

If you’re excited about building practical AI systems, solving difficult real-time problems, and helping shape what we build next, we’d love to hear from you.

Pay: $130,000.00 - $180,000.00 per year

Benefits:

• Dental insurance
• Health insurance
• Paid time off

Ability to Commute:

• Scottsdale, AZ 85260 (Required)

Work Location: In person