Research Fellow (Artificial Intelligence)

What matters most

• Combine AI with formal methods

  The role calls for research combining AI and formal methods, especially program synthesis, to improve AI-model explainability and robustness.

• Improve requirements engineering and testing

  Apply research approaches to AUMOVIO-NTU’s software engineering process, particularly requirements engineering and testing. The posting also calls for research experience in automated testing.

• Integrate research into development

  Develop and integrate AI algorithms into ongoing development, working with the development team and interacting with AUMOVIO-NTU business areas.

• Produce research and project outputs

  Prepare research papers and patent applications, and contribute to quarterly and annual reports and presentations. Communication and report-writing are stated requirements.

Likely challenges to think through

• Balance model properties with practical use

  Likely preparation challenge, not a confirmed project issue: an approach that improves explainability or robustness may be difficult to integrate into development. Prepare to explain how you would evaluate those properties alongside implementation effort and performance.

• Translate business needs into testable work

  Likely preparation challenge: business-area needs may not arrive as precise technical requirements. Prepare an example of turning a stated need into a requirement and testing approach while checking assumptions with the development team.

• Balance research outputs with development contributions

  Likely preparation challenge: papers, patent applications and periodic reports require different forms of communication from algorithm integration. Prepare to describe how you would document research results and development progress clearly for each audience.

Topics to prepare

• AI and formal methods

  The posting explicitly calls for research combining these areas. To revise, sketch how a formal property could constrain an AI-based approach, and identify what the property would and would not establish.

• Program synthesis

  Program synthesis is an explicit research requirement. To revise, work through a small example from specification to synthesized program, then explain how you would assess whether the result satisfies the specification.

• Explainability and robustness

  The stated research aim is to improve both properties of AI models using AI and formal-methods techniques. To revise, define a measurable criterion for each and outline an experiment that could reveal a trade-off.

• Requirements engineering and testing

  The role applies research to software engineering, especially requirements engineering and testing; automatic testing is also a stated background requirement. To revise, turn an ambiguous requirement into testable acceptance conditions and explain how those conditions support automated tests.

Skills to revise

• Python research prototyping

  Python is specifically emphasized, and the role develops AI algorithms. As a preparation exercise, implement a small algorithm in Python, document its inputs and outputs, and run a reproducible comparison against a simple baseline.

• Automated test construction

  Automatic testing is an explicit requirement. Practise writing executable tests for the Python prototype, including boundary cases and a failure case derived from a stated requirement.

• Development handoff

  The role involves integrating AI algorithms into ongoing development and working with the development team. As a suggested exercise, prepare a brief handoff for the prototype covering its interface, assumptions, tests and known limitations; no particular integration tool is specified.

• Research and technical reporting

  The posting calls for research papers, patent applications, quarterly and annual reports, and presentations. Practise summarizing a research result as a short paper-style abstract and a separate progress update that states evidence, limitations and next steps.

• Business-area communication

  The role includes interaction with AUMOVIO-NTU business areas. As a suggested exercise, explain the prototype's potential software-engineering use, constraints and open questions in a brief, non-specialist presentation.

• Select evidence for the stated requirements

  Prepare a brief evidence sheet covering the required PhD field and relevant research in AI, software engineering, automated testing and program synthesis. Identify publications, code or results you can discuss without sharing restricted material.

• Build a research project walkthrough

  Choose one project and prepare a clear account of its research question, method, evaluation and limitations. If relevant, explain how AI and formal methods or program synthesis affected model explainability or robustness—the combination this role investigates.

• Work through a requirements-and-testing design exercise

  As interview preparation, sketch how you would turn a sample software requirement into testable properties, then evaluate whether an AI-assisted approach improves the process. State assumptions and failure cases; the posting does not specify a particular system or toolchain.

• Prepare a Python integration example

  Be ready to explain a Python implementation you contributed to, including interfaces, tests and how you assessed its behavior. Relate it to the advertised responsibility to develop and integrate AI algorithms into ongoing development.

• Rehearse research and business-facing communication

  Prepare a short explanation of one research result for developers and a separate plain-language version for a business audience. Bring examples of research writing or reports you can share; the role includes papers, patent applications, periodic reports and presentations.

• Draft questions about the lab’s work

  Ask which requirements-engineering or testing problems are priorities, how researchers collaborate with the development team and business areas, and how research outcomes are evaluated for use in AUMOVIO’s development work.