H-1B Visa ML Engineer Jobs
ML Engineer roles qualify as H-1B visa specialty occupations because they require at least a bachelor's degree in computer science, statistics, or a related field. Most ML positions fall under SOC code 15-2051 (Data Scientists) or 15-1256 (Software Developers), which affects your prevailing wage level and LCA filing. Cap-subject H-1B registration opens each March for an October 1 start date.
Find H-1B Visa ML Engineer JobsOverview
Showing 5 of 2,655+ ML Engineer jobs
See all 2,655+ ML Engineer Jobs
Sign up for free to unlock all listings, filter by visa type, and get alerts for new ML Engineer roles.
Get Access To All JobsOur Mission
Rebuild how the world works, to make institutions work better for the people they serve.
About Brain Co.
Brain Co. builds AI-native operating systems for large, regulated institutions. Each system is built for a specific industry, powered by agents that push real workflows forward. Underneath it all is Atlas, our proprietary platform that keeps customers in control, secure by design, and never locked into one model.
Why Now
Brain Co. is entering its next phase of production deployments on a national scale with an elite team built from Palantir, Google, Meta, and Nvidia, and a growing footprint across government, insurance, health, and financial services.
Joining now means shaping both the company and a new category of applied AI. Every project here ships to production and is expected to create measurable customer value and impact.
You'll work alongside exceptional peers on some of the hardest problems in applied AI. It’s the kind of work you'll still be proud of in ten years from now.
Machine Learning Engineer, Platform
About the Role
So much of the work society depends on is still slower and harder than it should be. Permits take months. Claims sit unresolved. And AI hasn't changed that — because the bottleneck isn't the models. It's the institutional context AI needs to do the work: rules, history, relationships, and judgment scattered across people, documents, and legacy systems.
BrainCo exists to fix that. We build agent-native operating systems for the institutions society depends on, and our products are the first of their kind in the world — we were the first, anywhere, to fully automate construction permitting, and we're now doing the same across insurance and other industries. There is no playbook here, because no one has built this before.
As a Machine Learning Engineer on Platform, you'll build the core ML capabilities every product we ship stands on — built once, shared everywhere. This is the leverage seat in the company: improve document extraction, and every vertical improves with it; strengthen the blueprint foundation model, and every construction workflow gets sharper; ship a better improvement loop, and every system we've ever deployed keeps getting better on its own.
Come help build Atlas — our platform which includes a foundation model for the world's construction documents, extraction agents that read everything from policy stacks to financial filings, a unified eval system across every use case, model routing that puts the right model on the right task at the right cost. You'll own each capability end-to-end — from the pod that needs it this quarter to the abstraction that serves ten pods next year. Your customers are never abstract: they're the project pods building on your work, and through them, every institution we serve.
Who We're Looking For
You understand how machine learning actually works — not just the tooling, but the philosophy underneath: what a loss function really optimizes, how generalization breaks under distribution shift, why evaluation is where systems quietly go wrong. And you live at the bleeding edge of modern AI, with hard-won instincts for squeezing the most out of LLMs and agentic systems — prompting, fine-tuning, tool use, and reasoning. That combination is the job: you know when a fine-tuned segmentation model beats a VLM, when a rule engine beats both, and how to compose all three into a system more accurate than any single model. You treat frontier models as components to be measured, pushed, and engineered — never as magic.
You also have the platform instinct: you spot the general capability hiding inside three teams' specific requests — and know when generalizing is premature. You treat internal teams as real customers with real deadlines, and measure your success in their velocity.
Most of all, you're energized by building things that have never existed, and comfortable when the problem, the data, and the definition of success all have to be invented at once.
The Problems You'll Work On
Agents as shared capabilities. Document extraction, financial reporting, market data — built once, composed into many products. The dual bar: general enough for any pod to pick up, precise enough for decisions institutions stake their processes on.
Institutional Intelligence that compounds. Every verified correction improves the system twice: the corrected fact percolates to every application, and the system that builds the intelligence learns to build it better. You'll build the models behind both loops.
A foundation model for construction documents. The documents the built world runs on have never had a foundation model of their own. We have the data, the deployments, and the feedback loops to build one.
Model routing across every use case. The right model, at the right cost and latency, for every task — swapping frontier models underneath production systems without breaking institutional-grade guarantees.
One eval system for everything. A common language for quality across every use case — from segmentation models checking blueprints to agents adjudicating claims — that catches regressions before customers ever see them.
Composite AI systems and credit assignment. When a pipeline of vision models, VLM reasoning, and rule engines is wrong, which component failed? Because the components are shared, this is a platform problem — and one of the most interesting open problems in applied ML.
Continuous improvement, engineered. We promise customers their system gets measurably better every month it runs. You'll build the machinery that keeps that promise: capturing production corrections, triaging failures to the component that caused them, and turning that signal into retraining and safe redeployment — automatically, across every use case.
In This Role, You Will:
Turn pod needs into platform capabilities — find the general capability inside one team's specific, urgent request, without over-abstracting before the pattern is proven.
Own capabilities end-to-end. There is no handoff: whoever builds the capability owns its behavior in production, across every deployment that uses it.
Work at the research frontier with production stakes, turning LLMs, RL fine-tuning, and agentic systems into capabilities that dozens of institutional workflows depend on at once.
Serve customers on both sides of the wall — project pods as true customers, and when needed, the domain experts whose decisions your capabilities ultimately power.
Engineer for production reality, navigating accuracy, latency, cost, and reliability across environments far messier than any benchmark.
Raise the bar across the company. The platform is how learnings travel: what one pod discovers, you turn into something every pod inherits.
Our Mission
Rebuild how the world works, to make institutions work better for the people they serve.
About Brain Co.
Brain Co. builds AI-native operating systems for large, regulated institutions. Each system is built for a specific industry, powered by agents that push real workflows forward. Underneath it all is Atlas, our proprietary platform that keeps customers in control, secure by design, and never locked into one model.
Why Now
Brain Co. is entering its next phase of production deployments on a national scale with an elite team built from Palantir, Google, Meta, and Nvidia, and a growing footprint across government, insurance, health, and financial services.
Joining now means shaping both the company and a new category of applied AI. Every project here ships to production and is expected to create measurable customer value and impact.
You'll work alongside exceptional peers on some of the hardest problems in applied AI. It’s the kind of work you'll still be proud of in ten years from now.
Machine Learning Engineer, Platform
About the Role
So much of the work society depends on is still slower and harder than it should be. Permits take months. Claims sit unresolved. And AI hasn't changed that — because the bottleneck isn't the models. It's the institutional context AI needs to do the work: rules, history, relationships, and judgment scattered across people, documents, and legacy systems.
BrainCo exists to fix that. We build agent-native operating systems for the institutions society depends on, and our products are the first of their kind in the world — we were the first, anywhere, to fully automate construction permitting, and we're now doing the same across insurance and other industries. There is no playbook here, because no one has built this before.
As a Machine Learning Engineer on Platform, you'll build the core ML capabilities every product we ship stands on — built once, shared everywhere. This is the leverage seat in the company: improve document extraction, and every vertical improves with it; strengthen the blueprint foundation model, and every construction workflow gets sharper; ship a better improvement loop, and every system we've ever deployed keeps getting better on its own.
Come help build Atlas — our platform which includes a foundation model for the world's construction documents, extraction agents that read everything from policy stacks to financial filings, a unified eval system across every use case, model routing that puts the right model on the right task at the right cost. You'll own each capability end-to-end — from the pod that needs it this quarter to the abstraction that serves ten pods next year. Your customers are never abstract: they're the project pods building on your work, and through them, every institution we serve.
Who We're Looking For
You understand how machine learning actually works — not just the tooling, but the philosophy underneath: what a loss function really optimizes, how generalization breaks under distribution shift, why evaluation is where systems quietly go wrong. And you live at the bleeding edge of modern AI, with hard-won instincts for squeezing the most out of LLMs and agentic systems — prompting, fine-tuning, tool use, and reasoning. That combination is the job: you know when a fine-tuned segmentation model beats a VLM, when a rule engine beats both, and how to compose all three into a system more accurate than any single model. You treat frontier models as components to be measured, pushed, and engineered — never as magic.
You also have the platform instinct: you spot the general capability hiding inside three teams' specific requests — and know when generalizing is premature. You treat internal teams as real customers with real deadlines, and measure your success in their velocity.
Most of all, you're energized by building things that have never existed, and comfortable when the problem, the data, and the definition of success all have to be invented at once.
The Problems You'll Work On
Agents as shared capabilities. Document extraction, financial reporting, market data — built once, composed into many products. The dual bar: general enough for any pod to pick up, precise enough for decisions institutions stake their processes on.
Institutional Intelligence that compounds. Every verified correction improves the system twice: the corrected fact percolates to every application, and the system that builds the intelligence learns to build it better. You'll build the models behind both loops.
A foundation model for construction documents. The documents the built world runs on have never had a foundation model of their own. We have the data, the deployments, and the feedback loops to build one.
Model routing across every use case. The right model, at the right cost and latency, for every task — swapping frontier models underneath production systems without breaking institutional-grade guarantees.
One eval system for everything. A common language for quality across every use case — from segmentation models checking blueprints to agents adjudicating claims — that catches regressions before customers ever see them.
Composite AI systems and credit assignment. When a pipeline of vision models, VLM reasoning, and rule engines is wrong, which component failed? Because the components are shared, this is a platform problem — and one of the most interesting open problems in applied ML.
Continuous improvement, engineered. We promise customers their system gets measurably better every month it runs. You'll build the machinery that keeps that promise: capturing production corrections, triaging failures to the component that caused them, and turning that signal into retraining and safe redeployment — automatically, across every use case.
In This Role, You Will:
Turn pod needs into platform capabilities — find the general capability inside one team's specific, urgent request, without over-abstracting before the pattern is proven.
Own capabilities end-to-end. There is no handoff: whoever builds the capability owns its behavior in production, across every deployment that uses it.
Work at the research frontier with production stakes, turning LLMs, RL fine-tuning, and agentic systems into capabilities that dozens of institutional workflows depend on at once.
Serve customers on both sides of the wall — project pods as true customers, and when needed, the domain experts whose decisions your capabilities ultimately power.
Engineer for production reality, navigating accuracy, latency, cost, and reliability across environments far messier than any benchmark.
Raise the bar across the company. The platform is how learnings travel: what one pod discovers, you turn into something every pod inherits.
See all 2,655+ H-1B Visa ML Engineer Jobs
Sign up for free to unlock all listings, filter by visa type, and get alerts for new H-1B Visa ML Engineer Jobs.
Get Access To All JobsTips for Finding H-1B Visa Sponsorship as a ML Engineer
Map your degree to SOC codes
ML Engineer roles get filed under different SOC codes depending on the employer, most use 15-2051 or 15-1256. Pull your target SOC from O*NET before interviews so you can confirm your degree field aligns and anticipate any specialty occupation scrutiny.
Verify prevailing wage before negotiating
Use the OFLC Wage Search to look up the Level I through Level IV wage for your SOC code and work location before you accept an offer. Your employer's LCA must certify a wage at or above that threshold, and it affects what they can legally pay you.
Target employers already filing LCAs for ML roles
Search Migrate Mate to filter for employers with verified H-1B LCA filing history specifically for ML and data science roles. That filing record tells you the employer knows the process and has done it for your job type before.
Ask whether the employer is cap-exempt
Universities, nonprofit research institutions, and certain government-affiliated entities are cap-exempt, meaning your H-1B can start any time of year without waiting for the October 1 cap date. If you're open to those employers, your timeline compresses significantly.
Request premium processing if your start date is tight
USCIS offers premium processing on I-129 petitions, which guarantees a 15-business-day adjudication window. If you're transferring from OPT or a prior H-1B and your authorization gap is narrow, confirm early whether your employer will pay for it.
Document project scope and model ownership clearly
USCIS RFEs for ML Engineer petitions frequently challenge whether the role truly requires a specialized degree rather than general coding skills. Collect evidence showing you own model architecture decisions, not just implementation tasks, to support the specialty occupation case.
H-1B Visa ML Engineer: Frequently Asked Questions
Does an ML Engineer role qualify as a specialty occupation for H-1B purposes?
Yes, provided the position requires at least a bachelor's degree in a directly related field such as computer science, mathematics, or statistics. Roles focused on model research, algorithm design, or applied machine learning typically satisfy the specialty occupation standard. Roles that only require general programming skills with no field-specific degree requirement can face RFEs, so the job description needs to be precise about degree requirements.
Which SOC code do employers typically use when filing H-1B petitions for ML Engineers?
Most employers file under 15-2051 (Data Scientists) or 15-1256 (Software Developers and Software Quality Assurance Analysts). The choice affects the prevailing wage level your employer must certify on the LCA. You can look up both codes on O*NET to see which job zone and education requirements best match your role's actual duties before your employer submits the LCA to DOL.
How do I find employers who have actually sponsored H-1B visas for ML Engineer roles before?
Migrate Mate surfaces employers with verified DOL Labor Condition Application filing history filtered by occupation code, so you can see which companies have filed LCAs for ML and data science roles specifically. Checking LCA history is more reliable than asking recruiters, because it shows actual filings rather than stated willingness to sponsor.
Can I switch ML Engineer jobs while on H-1B status without losing my place in line?
Yes. Under AC21 portability, if your I-140 has been approved for at least 180 days and your new role is in the same or a substantially similar occupational classification, you can port your priority date to the new employer's petition. Your new employer still needs to file a new H-1B transfer petition before your current authorization ends, and you can start the new job once that petition is received by USCIS.
What documentation should I prepare before an employer files my H-1B petition as an ML Engineer?
You'll need your academic transcripts and degree certificates, any credential evaluations if your degree is from outside the U.S., a current resume detailing your project responsibilities, and any publications or patents that support your specialty. Your employer's attorney will also need your passport, current I-94 record, and prior immigration documents. Having these organized before the attorney engagement starts reduces delays significantly.