We're looking for a full-time, fully remote Virtual AI Engineer at $145,000 per year, open to applicants anywhere.
What you'll bring
You'll need a bachelor's degree, usually in computer science, data science, or a related field, along with hands-on experience building and deploying machine learning models. This role asks for 24 months of that experience specifically. Strong Python skills are a given, and you should be comfortable with at least one major deep learning framework already, since there won't be much time built in to learn PyTorch or TensorFlow from scratch on the job.
We're less concerned about which framework you know best than whether you've actually shipped something. A model that never left a notebook doesn't tell us much about how you'll handle the messier parts of this role.
The two years doesn't have to come from a single job with "AI Engineer" in the title. A backend engineer who spent that stretch increasingly focused on ML integration, or a data scientist who moved toward the deployment side of projects, both fit the bar as long as the hands-on production experience is genuinely there.
What the work actually looks like
- Design, train, and deploy machine learning models into production systems
- Optimize model performance once it's live, not just during development
- Integrate AI features directly into applications people actually use
Here's a good example from a few months back: a support-ticket classifier tested well and shipped without issue, but response times crept up noticeably during peak hours after launch. It turned out the model was being reloaded from disk on every single request instead of staying cached in memory. An easy fix once found, but it took someone who understood both the ML side and the serving infrastructure to catch it, since the model's predictions themselves were perfectly fine the whole time.
You'll also work closely with data scientists and software engineers to move models out of the experimentation phase and into a state reliable enough to run at scale. That handoff point is where a lot of the real engineering happens.
REST APIs come up constantly, since most AI features here get exposed to the rest of the product through one. Getting the interface right, with sensible error handling and reasonable response times, matters just as much as the model behind it, and a poorly designed API can make even a strong model feel unreliable to the teams consuming it.
Skills
- Python
- TensorFlow and PyTorch
- Machine learning pipelines
- MLOps
- REST APIs
- Cloud platforms, including AWS, GCP, and Azure
- Model deployment
- SQL
Deep familiarity with both TensorFlow and PyTorch isn't required. Most engineers here lean heavily on one and can read the other well enough to work with existing code, which is a realistic expectation rather than mastery of both.
Pay and what comes with it
The role pays $145,000 a year. On the benefits side, you're covered with retirement matching, time off, and health insurance through the employer, plus a real budget for conferences, courses, or GPU compute resources when a project needs more horsepower than the standard setup provides. Remote-work flexibility is built into the role by default, not treated as a special accommodation.
Naukri Mitra is running the hiring process here, and the GPU compute budget specifically has come in handy for engineers who wanted to prototype something heavier than their laptop could reasonably handle before bringing it to the team for review.
Raises are reviewed annually and reflect what someone's actually shipped, rather than time served alone. Taking a model from a rough prototype to something that reliably runs in production tends to carry real weight in that conversation, more so than incremental improvements to something already stable.
If you've been comparing AI engineer remote salary numbers, this one lands well above the median for someone with two years of hands-on ML deployment experience, which tracks with how competitive hiring's gotten for engineers who can actually ship models rather than just build them.
Day to day on the team
This is a small AI engineering group embedded inside a larger product organization, so you'll spend real time talking to people outside the ML world, not just other engineers. A product manager might ask why a feature's accuracy dropped after a recent update, and being able to explain that clearly, without drowning them in jargon, matters as much as fixing the underlying issue.
Among AI engineering jobs worldwide, the split between research-heavy roles and production-focused ones varies widely by company. This one sits firmly on the production side. If most of what excites you is reading papers and running experiments with no deployment attached, this probably isn't the right fit, but if you like seeing a model actually affect something a real user experiences, it's a good match.
Deployment work happens in small, frequent releases rather than big infrequent ones. A model update usually ships to a limited slice of traffic first, is closely watched for a few days, and then rolls out more widely once the numbers hold up.
How to apply
Send a resume along with a short note on a machine learning model you've deployed to production, including one thing that broke or surprised you after launch. Interviews include a technical round on ML fundamentals and deployment practices, followed by a shorter conversation about working across product and engineering teams.
