OPT Machine Learning Research Jobs
Machine Learning Research jobs on OPT require STEM designation, which qualifies F-1 students for a 24-month OPT extension beyond the standard 12 months. Most roles sit inside research labs or AI teams at universities, national laboratories, and technology companies, all of which routinely sponsor H-1B visa and O-1 visas for strong candidates.
Find OPT Machine Learning Research JobsOverview
Showing 5 of 517+ Machine Learning Research jobs
See all 517+ Machine Learning Research Jobs
Sign up for free to unlock all listings, filter by visa type, and get alerts for new Machine Learning Research roles.
Get Access To All JobsINTRODUCTION
The AI & ML Operations Engineer designs, builds, and maintains the engineering processes and platform capabilities needed to deploy, monitor, govern, and maintain deployed AI/ML solutions at scale. The role converts validated analytical work into reliable production solutions and manages the operational lifecycle of models and related services while meeting enterprise privacy, security, and support standards.
The engineer partners with data scientists, data engineers, central cloud and infrastructure teams, API and Interoperability teams, Cybersecurity, service owners, and relevant engineering teams to deliver maintainable solutions. The role remains hands-on in engineering, testing, deployment, and troubleshooting while using enterprise-managed capabilities for production data movement, integration, and infrastructure. Scientific method validation remains with the Data Scientist and designated reviewers; this position implements the operational controls for deployed solutions.
This role is part of a new AI & Automation engineering team and will directly design, build, deploy, and support production AI applications and services. Work may include agents, retrieval-augmented generation, orchestration, APIs and integrations, automated evaluation and testing frameworks, observability, monitoring, release and rollback processes, and model/application lifecycle management.
Responsibilities
Model deployment and orchestration. Deploy batch and real-time models and related AI services using approved CI/CD, containers, APIs, and workflow patterns. Build and maintain assigned model deployment and runtime components and coordinate dependencies with enterprise platform owners.
Release management. Automate packaging, versioning, testing, promotion, rollback, and recovery for models and services. Implement retraining or recalibration workflows using criteria defined with data scientists and the required validation and release approvals.
ML platform capabilities. Implement, integrate, configure, and operate approved feature stores, model registries, experiment tracking, and related ML lifecycle capabilities with enterprise platform owners. Partner with Cloud & AI Deployment Engineering and central cloud teams on compute, storage, networking, and deployment.
Deployed-model governance. Implement approved lifecycle controls, access restrictions, lineage, audit trails, change records, and release approvals for deployed models. Maintain traceability across data, code, model, and configuration versions.
Monitoring and reliability. Implement logging, alerts, and monitoring for model performance, drift, data quality, latency, availability, and errors. Work with data scientists, service owners, and relevant engineering teams on measures, thresholds, incident response, and service objectives.
Data and API integration. Work with central data-engineering teams on managed data pipelines and with enterprise API and Interoperability teams on real-time and application integrations, including MuleSoft or other approved API capabilities. Coordinate monitoring, recovery checkpoints, restart procedures, security, and support ownership.
Enterprise support and maintainability. Ensure production solutions have documented dependencies, operational procedures, escalation paths, and ongoing ownership before release. Arrange support with the responsible enterprise teams, including 24/7 coverage where required for critical services, rather than relying on an individual project owner.
AI application operations. Operate assigned LLM, retrieval-based, and agent applications using automated quality checks, model/tool tracing, prompt and configuration versioning, and usage/cost monitoring within the common production lifecycle.
Security and engineering standards. Apply enterprise privacy, security, encryption, role-based access, and sensitive-data controls. Follow common infrastructure-as-code, pull-request, and deployment practices with the responsible teams; bring needed tooling or design exceptions through established review processes.
Collaboration and improvement. Help scientists and partner teams move prototypes into supported production solutions. Review code, tests, and documentation, resolve incidents across dependencies, provide constructive technical feedback, and improve reliability, reproducibility, performance, and efficiency.
BASIC QUALIFICATIONS
- B.S./B.A. in Computer Science, Engineering, Information Systems, or a related field, or an equivalent combination of education and relevant experience.
- At least three years of experience in software engineering, DevOps, or MLOps, including work with production systems.
- Proficiency in Python or another modern programming language, such as Java or Go, sufficient to build, test, troubleshoot, and maintain production components.
- Hands-on experience deploying and operating batch and/or real-time machine-learning models in production.
- Experience with CI/CD tools such as GitHub Actions, GitLab CI, Jenkins, or Azure DevOps; source control, code review, automated testing, and controlled releases.
- Hands-on experience with containerization and orchestration technologies using enterprise-approved platforms and tools.
- Experience in cloud and/or hybrid environments such as AWS, Azure, or GCP.
- Working knowledge of SQL and experience with large datasets and data-workflow orchestration; ability to integrate model services with enterprise-managed pipelines and APIs.
- Experience implementing logging, metrics, monitoring, alerts, and operational diagnostics and investigating failures across model, application, data, and infrastructure dependencies.
- Understanding of security best practices, role-based access control, secrets management, and handling sensitive data.
- Ability to independently deliver assigned engineering work, communicate technical tradeoffs, and collaborate with data scientists, clinicians, data/API teams, cloud/DevOps, security, and support functions.
PREFERRED QUALIFICATIONS
- Experience in healthcare, health technology, or another regulated industry handling PHI/PII.
- Familiarity with EHR/EMR integration and healthcare data standards such as HL7/FHIR, ICD, CPT, LOINC, or SNOMED.
- Experience with ML lifecycle tools such as MLflow, Kubeflow, SageMaker, Vertex AI, or Azure ML, including model registries, feature stores, and experiment tracking.
- Experience with ETL/ELT, data warehouse or lakehouse platforms, enterprise API platforms, infrastructure as code, and formal production-support practices.
- Experience monitoring model drift, fairness, and explainability and implementing healthcare AI governance controls.
- Experience operating LLM, retrieval-based, or agent applications, including quality evaluation, tracing, prompt/configuration management, and cost monitoring.
COMPENSATION
- $109,449.60 - $154,523.20 Annual
INTRODUCTION
The AI & ML Operations Engineer designs, builds, and maintains the engineering processes and platform capabilities needed to deploy, monitor, govern, and maintain deployed AI/ML solutions at scale. The role converts validated analytical work into reliable production solutions and manages the operational lifecycle of models and related services while meeting enterprise privacy, security, and support standards.
The engineer partners with data scientists, data engineers, central cloud and infrastructure teams, API and Interoperability teams, Cybersecurity, service owners, and relevant engineering teams to deliver maintainable solutions. The role remains hands-on in engineering, testing, deployment, and troubleshooting while using enterprise-managed capabilities for production data movement, integration, and infrastructure. Scientific method validation remains with the Data Scientist and designated reviewers; this position implements the operational controls for deployed solutions.
This role is part of a new AI & Automation engineering team and will directly design, build, deploy, and support production AI applications and services. Work may include agents, retrieval-augmented generation, orchestration, APIs and integrations, automated evaluation and testing frameworks, observability, monitoring, release and rollback processes, and model/application lifecycle management.
Responsibilities
Model deployment and orchestration. Deploy batch and real-time models and related AI services using approved CI/CD, containers, APIs, and workflow patterns. Build and maintain assigned model deployment and runtime components and coordinate dependencies with enterprise platform owners.
Release management. Automate packaging, versioning, testing, promotion, rollback, and recovery for models and services. Implement retraining or recalibration workflows using criteria defined with data scientists and the required validation and release approvals.
ML platform capabilities. Implement, integrate, configure, and operate approved feature stores, model registries, experiment tracking, and related ML lifecycle capabilities with enterprise platform owners. Partner with Cloud & AI Deployment Engineering and central cloud teams on compute, storage, networking, and deployment.
Deployed-model governance. Implement approved lifecycle controls, access restrictions, lineage, audit trails, change records, and release approvals for deployed models. Maintain traceability across data, code, model, and configuration versions.
Monitoring and reliability. Implement logging, alerts, and monitoring for model performance, drift, data quality, latency, availability, and errors. Work with data scientists, service owners, and relevant engineering teams on measures, thresholds, incident response, and service objectives.
Data and API integration. Work with central data-engineering teams on managed data pipelines and with enterprise API and Interoperability teams on real-time and application integrations, including MuleSoft or other approved API capabilities. Coordinate monitoring, recovery checkpoints, restart procedures, security, and support ownership.
Enterprise support and maintainability. Ensure production solutions have documented dependencies, operational procedures, escalation paths, and ongoing ownership before release. Arrange support with the responsible enterprise teams, including 24/7 coverage where required for critical services, rather than relying on an individual project owner.
AI application operations. Operate assigned LLM, retrieval-based, and agent applications using automated quality checks, model/tool tracing, prompt and configuration versioning, and usage/cost monitoring within the common production lifecycle.
Security and engineering standards. Apply enterprise privacy, security, encryption, role-based access, and sensitive-data controls. Follow common infrastructure-as-code, pull-request, and deployment practices with the responsible teams; bring needed tooling or design exceptions through established review processes.
Collaboration and improvement. Help scientists and partner teams move prototypes into supported production solutions. Review code, tests, and documentation, resolve incidents across dependencies, provide constructive technical feedback, and improve reliability, reproducibility, performance, and efficiency.
BASIC QUALIFICATIONS
- B.S./B.A. in Computer Science, Engineering, Information Systems, or a related field, or an equivalent combination of education and relevant experience.
- At least three years of experience in software engineering, DevOps, or MLOps, including work with production systems.
- Proficiency in Python or another modern programming language, such as Java or Go, sufficient to build, test, troubleshoot, and maintain production components.
- Hands-on experience deploying and operating batch and/or real-time machine-learning models in production.
- Experience with CI/CD tools such as GitHub Actions, GitLab CI, Jenkins, or Azure DevOps; source control, code review, automated testing, and controlled releases.
- Hands-on experience with containerization and orchestration technologies using enterprise-approved platforms and tools.
- Experience in cloud and/or hybrid environments such as AWS, Azure, or GCP.
- Working knowledge of SQL and experience with large datasets and data-workflow orchestration; ability to integrate model services with enterprise-managed pipelines and APIs.
- Experience implementing logging, metrics, monitoring, alerts, and operational diagnostics and investigating failures across model, application, data, and infrastructure dependencies.
- Understanding of security best practices, role-based access control, secrets management, and handling sensitive data.
- Ability to independently deliver assigned engineering work, communicate technical tradeoffs, and collaborate with data scientists, clinicians, data/API teams, cloud/DevOps, security, and support functions.
PREFERRED QUALIFICATIONS
- Experience in healthcare, health technology, or another regulated industry handling PHI/PII.
- Familiarity with EHR/EMR integration and healthcare data standards such as HL7/FHIR, ICD, CPT, LOINC, or SNOMED.
- Experience with ML lifecycle tools such as MLflow, Kubeflow, SageMaker, Vertex AI, or Azure ML, including model registries, feature stores, and experiment tracking.
- Experience with ETL/ELT, data warehouse or lakehouse platforms, enterprise API platforms, infrastructure as code, and formal production-support practices.
- Experience monitoring model drift, fairness, and explainability and implementing healthcare AI governance controls.
- Experience operating LLM, retrieval-based, or agent applications, including quality evaluation, tracing, prompt/configuration management, and cost monitoring.
COMPENSATION
- $109,449.60 - $154,523.20 Annual
See all 517+ OPT Machine Learning Research Jobs
Sign up for free to unlock all listings, filter by visa type, and get alerts for new OPT Machine Learning Research Jobs.
Get Access To All JobsTips for Finding OPT Sponsorship in Machine Learning Research
Target STEM-designated programs
Verify your degree program carries an approved STEM CIP code before applying. Machine Learning Research roles qualify for the 24-month OPT extension only when your degree field matches the STEM designation list maintained by ICE.
Apply to research-heavy employers first
Universities, national laboratories, and corporate AI research divisions hire ML researchers regularly and have established immigration counsel on staff. These employers understand OPT timelines and are far more likely to sponsor H-1B visa or O-1 petitions after your authorization period.
File your OPT STEM extension early
USCIS recommends filing your 24-month STEM extension application 90 days before your initial OPT EAD expires. A timely filing gives you a 180-day automatic extension cap that keeps you authorized while USCIS processes your application.
Emphasize publications and research output
ML research hiring managers weight peer-reviewed publications, conference papers at NeurIPS, ICML, or ICLR, and open-source model contributions heavily. Strong research output also strengthens future O-1A or EB-1A immigration petitions based on extraordinary ability.
Clarify your authorization window upfront
Tell recruiters your exact EAD expiration date and STEM extension eligibility in the first conversation. Research teams plan hiring timelines months ahead, so knowing you have two to three years of work authorization removes a common sponsorship hesitation early in the process.
Build toward O-1 eligibility from day one
Machine Learning researchers accumulate O-1 qualifying evidence naturally through citations, peer review, judging competitions, and speaking at conferences. Documenting these activities from your first OPT role creates a strong immigration record independent of the H-1B lottery.
Machine Learning Research OPT: Frequently Asked Questions
Do Machine Learning Research jobs qualify for the 24-month OPT STEM extension?
Yes, provided your underlying degree is in a STEM-designated field such as Computer Science, Statistics, Mathematics, or Electrical Engineering. Machine Learning Research roles fall squarely within the STEM occupation categories that satisfy the extension criteria. You should confirm your specific CIP code with your DSO before filing.
How do I find Machine Learning Research employers that sponsor OPT and H-1B?
Migrate Mate filters job listings specifically for employers willing to sponsor international candidates on OPT, so you can focus your search on companies that have already committed to sponsoring work authorization rather than asking cold. Corporate research labs at large technology companies, university research groups, and federally funded AI institutes are the most active sponsors in this field.
Can I work as an independent or contract ML researcher on OPT?
Self-employment is not permitted on standard OPT. Your work must be directly related to your degree field and performed for an employer who reports your employment to your DSO within 10 days of your start date. Contract or consulting arrangements may be permissible if structured through an employer of record who fulfills the reporting requirement and the work is degree-related.
What happens to my OPT status if my Machine Learning Research position ends before my EAD expires?
OPT allows a cumulative unemployment limit of 90 days during the initial 12-month period and an additional 60 days during the 24-month STEM extension, for 150 days total. Exceeding these limits terminates your F-1 status. If your position ends, you should begin a new job search immediately and report any employment changes to your DSO without delay.
Does Machine Learning Research experience improve my long-term visa sponsorship prospects?
Substantially. Peer-reviewed publications, conference presentations at venues like NeurIPS or ICML, and measurable research impact build direct evidence for O-1A extraordinary ability petitions and EB-1A or EB-2 NIW green card categories. Employers in this field are also among the most experienced H-1B sponsors, meaning your transition from OPT to long-term status is a well-established path rather than an exception.