Job Openings › ML Evaluation & Data Systems Engineering — Master's/Doctorate-Level Postgraduate Internship

ML Evaluation & Data Systems Engineering - Master's/Doctorate-Level Postgraduate Internship

Apply Now

About Us

Vosyn is a Language Synthesis AI company based in Toronto. We build technology that removes language barriers, so people can access and understand information in any language. We are preparing for an IPO, and the people who join us now help shape both the product and the company itself. Good ideas get heard here, whoever they come from.

About the Role

We are seeking a motivated and technically capable master's- or doctorate-level recent graduate or final-year student to join our Hong Kong research and development team as an ML Evaluation & Data Systems Engineer. This role focuses on building the systems used to measure, validate, reproduce, and improve the performance of multilingual AI models and applications. You will help create evaluation frameworks, multilingual benchmarks, data pipelines, experiment-tracking systems, audit logs, and reporting tools that allow Vosyn to demonstrate how well its AI systems preserve meaning, context, tone, cultural relevance, and accuracy across languages. This is an important role within our AI development process. You will help ensure that model improvements are measurable, technical experiments are reproducible, data is traceable, and product-quality claims are supported by reliable evidence.

Key Responsibilities

  • Design and develop evaluation frameworks for multilingual large language models and localization systems.
  • Build automated and human-review workflows for measuring AI output quality.
  • Establish benchmarks for contextual accuracy, semantic preservation, cultural appropriateness, consistency, latency, hallucination, safety, and overall usability.
  • Develop multilingual test datasets and evaluation scenarios covering Cantonese, Mandarin, English, and other strategically relevant languages.
  • Create processes for dataset ingestion, cleaning, validation, annotation, version control, and quality assurance.
  • Develop data pipelines that support model evaluation, experimentation, reporting, and continuous improvement.
  • Implement experiment-tracking systems for model versions, prompts, configurations, datasets, outputs, and evaluation results.
  • Design audit-log structures that connect source content, model configuration, output, reviewer feedback, quality scores, and corrective actions.
  • Build dashboards and reports that communicate AI performance to technical teams, management, pilot partners, and other stakeholders.
  • Support the design of human-evaluation rubrics, reviewer instructions, quality standards, and disagreement-resolution processes.
  • Apply statistical methods to compare model versions and determine whether observed improvements are meaningful.
  • Monitor model and data quality and help identify performance degradation, inconsistencies, bias, or unexpected behaviour.
  • Collaborate with applied AI engineers to test model improvements and identify areas requiring further development.
  • Support reproducible model experimentation, validation, and pilot-readiness testing.
  • Document system architecture, evaluation methodologies, data lineage, technical decisions, results, and recommended improvements.
  • Maintain records that support product development, responsible AI practices, intellectual property documentation, and future enterprise requirements.

About You

  • Currently enrolled in or recently graduated from a master's or doctorate program in Computer Science, Artificial Intelligence, Machine Learning, Data Science, Data Engineering, Statistics, Software Engineering, Information Systems, or a related field. Enrollment in or completion of a master's or doctorate program is mandatory.
  • Strong programming ability in Python.
  • Proficiency in SQL and experience working with structured and unstructured datasets.
  • Strong understanding of machine learning, data pipelines, model evaluation, or analytical methods.
  • Familiarity with large language models, natural language processing, transformer-based systems, or generative AI.
  • Experience with model evaluation, experiment design, statistical testing, benchmarking, or human-evaluation methods is an asset.
  • Familiarity with databases, APIs, cloud environments, data visualization, and dashboard-development tools.
  • Experience with experiment-tracking or MLOps platforms is an asset.
  • Familiarity with tools such as MLflow, Weights & Biases, DVC, Airflow, Prefect, or comparable technologies is an asset.
  • Understanding of data lineage, model observability, data governance, responsible AI, or auditability is an asset.
  • Experience working with multilingual data or language-based AI systems is an asset.
  • Professional proficiency in Cantonese or Mandarin is a strong asset.
  • Strong analytical, organizational, and problem-solving skills.
  • Excellent attention to detail and a disciplined approach to documentation and version control.
  • Ability to translate technical findings into clear reports and recommendations.
  • Effective written and verbal communication skills.
  • Ability to work independently while collaborating with a distributed, multidisciplinary team.
  • Must be based in Hong Kong, or able to work from Hong Kong, and legally eligible to work in Hong Kong.

Exceptional talent often emerges from diverse paths. If you possess profound curiosity, a genuine passion for continuous personal and professional growth, and a strong desire to apply your abilities to create meaningful impact, we encourage you to apply even if your background does not align perfectly with every qualification.

At Vosyn, you have the opportunity to contribute to a growing global organisation shaped by diversity of thought, experience and cultures. Our interns come from leading universities around the world, bringing fresh perspectives to work that moves the company forward. We invite you to join us on this exciting journey, applying your abilities, learning through real contributions and bringing the best version of yourself to the team you will work alongside.

Why Join Us

  • Work inside a start-up as it prepares for an IPO, and see how that actually gets done.
  • Work directly with senior management and members of the advisory board.
  • Build experience you can point to: real projects with real outcomes.
  • Take ownership of work that matters to the company, with the pace and responsibility of a start-up.
Apply Now

DEI and Workplace Safety

Vosyn is an equal-opportunity employer. We hire people from different backgrounds, disciplines, and countries because the work is better for it, and we want every member of the team to feel respected and supported. We take workplace safety seriously, follow the regulations that apply to us, and expect everyone here to look out for one another. If that sounds like a place you would want to work, we encourage you to apply.

Recruitment Process

  1. 1Apply
  2. 21st Interview
  3. 32nd Interview
  4. 43rd Interview
  5. 5Group Discovery Session
  6. 6Next Step Form
  7. 7Contract
  8. 8Onboarding

Your placement is an opportunity to contribute at a richly rewarding stage of growth, working directly with senior management and members of the advisory board while gaining an inside view of how a company develops. You will take ownership of work on the company's critical path, building experience you can carry into your next chapter, with priority consideration for full-time opportunities after your internship. If academic commitments affect your availability, share them with us so we can explore what is possible within your program's placement requirements. Thank you for considering a journey with us. We look forward to learning what you could contribute and what you hope to discover.

Ready to join Vosyn as our next ML Evaluation & Data Systems Engineering — Master's/Doctorate-Level Postgraduate Internship?Apply Now

DISCLAIMER

CONFIDENTIALITY

The information contained in this presentation (the “Presentation”) is being provided on a confidential basis for informational and discussion purposes only. The information set forth herein does not purport to be complete or to contain all information that a recipient would deem relevant in analyzing Vosyn Inc. and VosynVerse HK Limited (together, the “Company”). The information contained herein must be treated in a confidential manner and may not be reproduced, used, or disclosed, in whole or in part, without the prior written consent of the Company. Disclosure to persons other than the recipient and its representatives, who themselves are bound by confidentiality restrictions, is prohibited.

Certain information contained in this presentation constitutes “forward-looking statements,” which can be identified by the use of forward-looking terminology such as “may,” “will,” “should,” “expect,” “anticipate,” “target,” “project,” “estimate,” “intend,” “outlook,” “continue,” or “believe,” or the negatives thereof or other variations thereon or comparable terminology. Forward-looking statements are necessarily based upon a number of estimates and assumptions that, while considered reasonable by management, are inherently subject to significant business, economic, and competitive uncertainties and contingencies.

Forward-looking statements involve known and unknown risks, uncertainties, and other factors that may cause actual financial results, performance, or achievements to be materially different from the estimated future results, performance, or achievements expressed or implied by those forward-looking statements, and the forward-looking statements are not guarantees of future performance. Except as required by law, we disclaim any obligation to update or revise any forward-looking statements, whether as a result of new information, events, or otherwise. No forward-looking statements or projections can be guaranteed. Accordingly, you should not place undue reliance on any forward-looking statements or information.

No representation or warranty, expressed or implied, is made as to the accuracy or completeness of the information contained in this presentation. Certain information contained herein has been secured from third-party sources believed to be reliable, but we make no representations or warranties as to the accuracy of such information and accept no liability therefor.

This presentation is not, and under no circumstances is to be construed as, a prospectus, offering memorandum, or an advertisement or a public offering of securities. Offers to sell, or solicitations of offers to buy, any security can only be made through official offering documents that contain important information about statutory rights of action and a description of risk factors.

FUTURE-ORIENTED FINANCIAL INFORMATION

The Presentation includes forward-looking information about prospective results of operations, financial position, and EBITDA, based on assumptions about future economic conditions and courses of action and that is not presented in the format of a historical balance sheet, income statement, or cash flow statement, including, but not limited to, the information contained on slides 8, 9, 10, and 12 in this Presentation (the “Financial Outlook”). Actual results may vary from the Financial Outlook summarized in this Presentation.

HISTORICAL INFORMATION ILLUSTRATIVE ONLY

Historical statements contained in this document regarding past trends or activities should not be taken as a representation that such trends or activities will continue in the future. In this regard, certain financial information contained herein has been extracted from, or based upon, information available in the public domain and/or provided by the Company. In particular, historical results, including with respect to other AI or technology companies, should not be taken as a representation that such trends will be replicated in the future with respect to an investment in the Company.

VALUATION AND SHARE-COUNT PRESENTATION

Investor-relations materials will display “Aggregated Share Value as of [Date]” and related share-count figures as of the same date. These figures may be estimates, may be unaudited, and do not reflect subsequent issuances, cancellations, conversions, or re-pricings unless expressly stated.

THIRD-PARTY REFERENCES AND “STRATEGIC PARTNERSHIP” FRAMING

References to third-party companies, platforms, or programs describe technology usage, interoperability, participation in standard startup or cloud-credit programs, or non-exclusive collaboration only, and do not imply equity relationships, exclusivity, revenue commitments, or endorsements unless expressly stated in an official Company announcement. Mentions of Google, Microsoft, and NVIDIA should be understood in this manner unless the Company has issued a public statement defining such relationships as strategic partnerships.

MARKETING CLAIMS AND SHAREHOLDER COUNTS

Any statement such as “we have raised millions of dollars from over 3,200 shareholders” must be accurate as of August 2025 and supported by Company records or public filings. Where precision is not available for public use, use date-stamped phrasing such as “thousands of shareholders (as of August 2025).” Claims referring to “industry leaders,” “strategic partners,” or “tech giants” must be consistent with the Third-Party References section and with published Company announcements.

NO ADVICE

This Teaser does not contain investment, legal, accounting, tax, or financial advice. Each recipient should conduct its own due diligence and consult qualified advisors.

OFFICIAL DOCUMENTS CONTROL

In the event of any inconsistency between this Teaser and the official offering documents, the official offering documents govern.

UPDATES

The Company undertakes no obligation to update or revise this Teaser except as required by law. Terms, pricing, incentives, timelines, and availability are subject to change without notice. References to prospective events or transactions — including any shareholder buybacks, IPOs, or global launches — are forward-looking in nature. Such items are illustrative of current plans and strategic intent only, and are subject to market conditions, board approval, regulatory clearance, and the successful execution of underlying business milestones. No assurance can be given that these events will occur as presently contemplated.