Machine Learning Engineer

Together for Talent; Pittsburgh (Allegheny County); On-site; USD 125,000 - 175,000; Full time

Job summary

Together for Talent is seeking a hands-on Machine Learning Engineer to own the computer‑vision models at the heart of an early-stage AI platform for real-time quality inspection. You will work with production-line imagery, build and deploy defect detection models, and optimize them for edge hardware.

The role emphasizes on-site work in Pittsburgh, PA, with 3+ years in ML engineering focused on computer vision, and strong Python/PyTorch or TensorFlow expertise.

Qualifications

• 3+ years of ML engineering with a strong computer-vision focus.
• Strong Python and PyTorch or TensorFlow experience.
• Experience training, evaluating, and deploying deep-learning models in production.
• Solid grounding in classical computer vision and image processing (OpenCV or similar).
• Familiarity with edge inference and model optimization: ONNX, TensorRT, CUDA.
• Comfortable working with messy real-world data in a startup environment.
• Willing to work on-site in Pittsburgh, PA.

Responsibilities

• Build, train, evaluate, and improve computer‑vision models for defect detection and classification
• Develop anomaly‑detection approaches for rare, changing, or poorly labeled defects
• Optimize and deploy models to edge hardware, balancing accuracy, latency, and throughput
• Work closely with hardware and embedded engineers across the inspection pipeline
• Improve data workflows: labeling, curation, augmentation, retraining, and production monitoring
• Diagnose production model issues and help establish ML tooling and best practices

Skills

Python

PyTorch

TensorFlow

Deep learning

Computer vision

OpenCV

CUDA

Latency optimization

Tools

ONNX

TensorRT

NVIDIA edge platforms

Job description

Compensation:

$125K–$175K base + early-stage equity

We’re partnering with an early-stage AI company building real-time quality inspection systems for manufacturers. Their platform combines high-resolution cameras, edge hardware, and machine learning to catch product defects as they happen—reducing waste, rework, and manual inspection.

They are looking for a hands‑on Machine Learning Engineer to own the computer‑vision models at the center of the platform. You’ll work with real production‑line imagery, develop defect and anomaly‑detection models, and deploy them on edge hardware where speed and reliability matter.

WHAT YOU’LL DO

• Build, train, evaluate, and improve computer‑vision models for defect detection and classification
• Develop anomaly‑detection approaches for rare, changing, or poorly labeled defects
• Optimize and deploy models to edge hardware, balancing accuracy, latency, and throughput
• Work closely with hardware and embedded engineers across the inspection pipeline
• Improve data workflows: labeling, curation, augmentation, retraining, and production monitoring
• Diagnose production model issues and help establish ML tooling and best practices

WHO YOU ARE

• 3+ years of ML engineering experience with a strong computer‑vision focus
• Strong Python and hands‑on PyTorch or TensorFlow experience
• Experience training, evaluating, and deploying deep‑learning models in production
• Solid grounding in classical computer vision and image processing, including OpenCV or similar tools
• Familiarity with edge inference and model optimization: ONNX, TensorRT, quantization, CUDA, NVIDIA GPUs, latency profiling, or related tools
• Comfortable working with messy real‑world data and moving quickly in a startup environment
• Willing to work on‑site in Pittsburgh, PA

PREFERRED EXPERIENCE

• Manufacturing, industrial inspection, machine vision, robotics, or other high‑throughput visual systems
• Visual anomaly detection, unsupervised/self‑supervised learning, or rare‑event classification
• Vision‑language models, zero‑/few‑shot classification, or multimodal AI
• AWS/GCP, NVIDIA edge platforms, CUDA, embedded AI, or startup experience

WHAT WE OFFER

• $125K–$175K base salary plus meaningful early‑stage equity
• A high‑ownership role on a growing, well‑funded technical team
• The opportunity to build ML systems that make an immediate impact in real manufacturing environments
• Relocation support and a collaborative, in‑person engineering culture
• Roughly 50‑hour workweeks on average, with no expected weekend work

If you enjoy taking computer‑vision models from experimentation to reliable, real‑world deployment, we’d love to hear from you.