Job Description
Our client is building vision agents for large venues such as hotels and casinos— powering real-time video analytics and intelligent surveillance across hundreds of camera streams. Our systems run on-prem in some of the largest resorts in Las Vegas, and many more in the pipeline.
They’re a highly technical team shipping deep tech into one of the most operationally demanding and dynamic environments.
The Role
We’re looking for a Forward Deployed ML Engineer who blends strong technical ML/CV ability with comfort deploying systems in the field.
You will own our real-time vision pipelines end-to-end and be the technical face of the client’s inside casinos.
This role is not a back-office research job.
You will:
- Ship models into production
- Debug production pipelines at client sites
- Build new ML features ranging from classical ML, computer vision and LLMs
- Work hands-on with GPU servers & multi-camera systems
- Collaborate with customer surveillance teams and distribution partners
If you love solving real-world problems in messy environments, this is your role.
What You’ll Do
Train, tune, and update/deploy deep learning models at client sites
Maintain low-latency inference pipelines on-premise using PyTorch, ONNX, and TensorRT and Triton.
Build training data processing pipelines, QA/QC labeling and coordinate work with our labelling teams
Work closely with customers and with the product manager to experiment and ship new features
2-3 years of experience in machine learning with strong knowledge about not just deep learning but also classical ML (You’re an ML engineer first — someone who can train models, tune them, debug them in the wild, and build the software around them to make them production-ready.).
Strong skills in Linux, Docker, and shipping models as services.
Comfortable working in live production environments with minimal supervision.
A startup mindset — resourceful, adaptable, and excited to work across ML, backend, and DevOps boundaries.
Nice to Have
Experience with GStreamer, FFmpeg, or RTSP (or similar protocol) video pipelines.
Experience with Triton Server, model optimization using TensorRT and other deep learning acceleration frameworks.
Work remotely Monday - Friday, 40 hours a week (no weekends)
Vacation: 10 business days a year
Holidays: 5 National Holidays a year
Company Holidays: 5 Company Holidays a year (Christmas Eve, Christmas Day, New Year’s Eve, New Year’s Day, Zipdev Day)
Parental Leave
Health Care Reimbursement
Active Lifestyle Reimbursement
Quarterly Home Office Reimbursement
Payroll Deduction Purchase Plans
Longevity Bonus
Continuous Learning Bonus
Access to Training and Professional Development Platforms
Did we mention it’s REMOTE?!!
One of our core values at Zipdev is “Be authentic.” that’s why we encourage you to answer the application form in your own words; we are interested in getting to know you, not a digital assistant.
Wondering how our remote environment or our payment method work? We’ve put together some helpful answers in our FAQs at the bottom our our career site. Take a look and let us know if you have any other questions!
