New opportunity

AI Software Developer

National University of Singapore · Singapore

About this role

The AI Software Developer at NUS Information Technology builds scalable applications and AI-powered features, using AI tools throughout the software development lifecycle. The role is based at Kent Ridge Campus.

Responsibilities

Use GitHub Copilot, Cursor and Claude for requirements analysis, design, coding, refactoring, debugging, testing, documentation, code review and DevSecOps automation.
Develop full-stack applications with C#/.NET Core, Node.js and Python, and React or Angular frontends. Integrate OpenAI and Anthropic LLM APIs, prompt engineering, semantic search and intelligent workflows.
Use AI tools to generate tests, identify bugs and improve code quality and security. Mentor colleagues on AI tool adoption, establish best practices and improve development productivity.

Requirements

7-12+ years of software development experience with proven full-stack delivery, plus a bachelor's degree in Computer Science or equivalent practical experience.
Expert use of AI coding assistants across the development lifecycle and experience integrating LLM APIs into applications with prompt engineering.
Experience with the stated backend and frontend stack; SQL databases such as Microsoft SQL Server, PostgreSQL or MySQL; NoSQL technologies such as MongoDB or Redis; RESTful APIs, OAuth2, JWT and Docker.
Strong clean-code and design-pattern fundamentals; unit, integration and end-to-end testing; CI/CD and Agile/SDLC experience; stakeholder communication and mentoring junior engineers.

NUS IT states that it offers competitive remuneration and flexible work arrangements, without specifying a salary or work-mode schedule.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

GitHub CopilotCursorClaudeC#.NET CoreNode.jsPythonFastAPIFlaskReactAngularTypeScriptLLM APIsOpenAI APIAnthropic APIPrompt engineeringSemantic searchSQLMicrosoft SQL ServerPostgreSQLMySQLMongoDBRedisRESTful API designOAuth2JWTDockerDevSecOpsCI/CDUnit testing

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

AI assistance across the SDLC
The posting calls for expert use of GitHub Copilot, Cursor and Claude from requirements analysis and design through coding, testing, documentation, review and DevSecOps automation.
Full-stack application delivery
Build scalable applications using C#/.NET Core, Node.js or Python with React or Angular frontends. The role also calls for RESTful APIs, databases, authentication and Docker.
AI-powered features
Integrate OpenAI and Anthropic APIs and implement prompt engineering, semantic search and intelligent workflows in applications.
Engineering quality and security
Use AI tools to generate tests, review code and identify bugs while maintaining clean-code standards, unit/integration/end-to-end testing, CI/CD and security.
Technical leadership
Mentor colleagues on AI-tool adoption, establish best practices, communicate with stakeholders and improve development productivity.

Likely challenges to think through

Balancing AI-assisted speed with review
Preparation hypothesis: Faster generated code may increase the need to verify correctness, security and maintainability. Prepare an example of how you reviewed AI output, tested it at multiple levels and decided what needed human revision.
Making LLM features dependable
Preparation hypothesis: Prompt engineering and semantic search can produce useful results without guaranteeing consistent answers. Prepare to explain how you would evaluate outputs, handle unsuitable responses and decide when a conventional application workflow is preferable.
Leading adoption without weakening standards
Preparation hypothesis: Encouraging AI-tool use while mentoring engineers may require balancing productivity gains against shared code-quality practices. Prepare a concrete approach to best practices, code review and measuring whether a workflow improvement helps.

Practice your answers

Technical practice

How would you use AI coding assistants from requirements analysis through deployment?

What to coverThe role calls for using GitHub Copilot, Cursor and Claude across the software development lifecycle. Preparation suggestion: explain where you would use them for design, coding and documentation, and how you would independently verify their outputs before release.

How would you design a scalable full-stack application using the stated stack?

What to coverThe posting names C#/.NET Core, Node.js and Python backends, with React or Angular frontends using TypeScript. Preparation suggestion: describe how you would divide responsibilities across the frontend, APIs and backend, and justify your choices for maintainability and scale.

How would you integrate an OpenAI or Anthropic API into an application?

What to coverIntegrating both providers' LLM APIs is part of the advertised work. Preparation suggestion: cover API boundaries, error handling, latency and cost trade-offs, and how you would test the feature without assuming model responses are always reliable.

How would you develop and evaluate prompts for an AI-powered feature?

What to coverThe role explicitly includes prompt engineering. Preparation suggestion: explain how you would define expected behavior, test representative inputs and failure cases, and check whether prompt changes improve results without introducing regressions.

How would you implement semantic search within an intelligent workflow?

What to coverSemantic search and intelligent workflows are named duties. Preparation suggestion: discuss document preparation, retrieval relevance and how you would evaluate whether retrieved content helps the workflow produce useful results.

How would you choose between the listed SQL and NoSQL technologies?

What to coverThe stack includes SQL databases and NoSQL technologies such as MongoDB and Redis. Preparation suggestion: compare data relationships, query needs and access patterns, then explain how you would validate the choice against an application's requirements.

How would you design and secure a RESTful API for the frontend?

What to coverThe posting specifies RESTful API design, OAuth2 and JWT. Preparation suggestion: address resource design, authentication and authorization boundaries, token validation and error responses without assuming a particular identity provider.

How would you test and debug an AI-assisted code change?

What to coverThe role calls for AI-generated tests, bug identification, code review and unit, integration and end-to-end testing. Preparation suggestion: explain how you would reproduce a defect, select the right test level and review generated tests for meaningful assertions and missed edge cases.

How would you build quality and security checks into delivery?

What to coverCI/CD, Docker and DevSecOps automation appear in the posting. Preparation suggestion: outline appropriate build, test and review checkpoints, including how you would inspect AI-generated changes rather than treating automated output as approval.

How would you refactor AI-generated code while preserving behavior?

What to coverThe role emphasizes clean code, design patterns, refactoring and high engineering standards. Preparation suggestion: describe how you would identify a maintainability problem, choose a proportionate design improvement and use tests and review to check the refactor.

Behavioural practice

Tell me about a time you helped colleagues adopt a development tool or practice.

What to coverMentoring colleagues on AI tool adoption and establishing best practices are stated duties. Preparation suggestion: use a real situation/action/result example showing how you supported adoption and assessed whether the change helped.

Describe a time you resolved unclear requirements with a stakeholder.

What to coverThe role includes requirements analysis and stakeholder communication. Preparation suggestion: use a real situation/action/result example explaining the ambiguity, the questions you asked and how the agreed scope affected delivery.

Tell me about a time you challenged a proposed code change on quality or security grounds.

What to coverCode review, bug identification, security improvement and high engineering standards are part of the role. Preparation suggestion: use a real situation/action/result example covering the risk you identified, how you communicated it and the outcome.

Describe a time you mentored a junior engineer through a technical problem.

What to coverThe posting specifically calls for mentoring junior engineers. Preparation suggestion: use a real situation/action/result example that distinguishes the guidance you gave from the engineer's own work and explains what changed.

Tell me about a time you improved a team's development workflow.

What to coverThe role calls for technical leadership and productivity improvements through AI-assisted workflows. Preparation suggestion: use a real situation/action/result example showing the workflow problem, your intervention and the evidence you used to judge its effect.

Sharpen your knowledge

Topics to prepare

AI-assisted software development lifecycle
The posting calls for expert use of GitHub Copilot, Cursor and Claude from requirements analysis through DevSecOps automation. Revise where assistant output needs human verification: take a sample feature request and map useful AI assistance, review checks and deliverables at each SDLC stage.
LLM feature design
The role explicitly includes OpenAI and Anthropic APIs, prompt engineering, semantic search and intelligent workflows. Sketch a small feature’s request flow, prompt inputs, search step and response checks; explain what each component contributes.
Full-stack architecture and data choices
The stated stack spans C#/.NET Core, Node.js, Python, React or Angular with TypeScript, SQL and NoSQL. Revise by drawing a RESTful feature across frontend, backend and storage, then justify a SQL versus MongoDB or Redis choice for its data.
API security and delivery
RESTful APIs, OAuth2, JWT, Docker, CI/CD and DevSecOps are named in the posting. Trace an authenticated request from client to API, noting where authorization checks, secrets handling, containerization and pipeline checks belong.
Engineering quality and technical leadership
The posting emphasizes clean code, design patterns, unit/integration/E2E testing, stakeholder communication and mentoring. Review a sample design for maintainability and test coverage, then prepare a brief explanation of its trade-offs for both an engineer and a stakeholder.

Skills to revise

Direct and verify AI coding assistants
Practice using GitHub Copilot, Cursor or Claude, where available, to turn a small requirement into code, tests and documentation. Inspect the changes yourself and record a bug or security concern you would check before accepting them; the posting expects AI-assisted development and review.
Implement a full-stack feature
As a revision exercise, build one small REST endpoint in a stated backend option—C#/.NET Core, Node.js or Python with FastAPI/Flask—and call it from React or Angular with TypeScript. Be ready to explain how you would work across the other named stack options; the role requires full-stack delivery.
Integrate an LLM-powered workflow
Prototype, or outline if API access is unavailable, a feature using an OpenAI or Anthropic API with a deliberate prompt and a semantic-search step. Walk through its inputs and outputs; these integrations are explicit duties.
Apply storage and authentication choices
Extend the sample API with a SQL-backed record and explain where MongoDB or Redis would be appropriate instead. Trace an OAuth2/JWT-authenticated request and check its authorization path; databases and both authentication technologies appear in the stack.
Test and package a change
Add unit, integration and E2E test cases to the sample feature; use an AI assistant to propose additional cases, then review them for gaps. Package the application with Docker and outline CI/CD checks, reflecting the posting’s testing, containerization and delivery expectations.
Explain and mentor on AI-tool adoption
Rehearse a short walkthrough for junior engineers showing one useful AI-assisted workflow, its review safeguards and how it improves a development task. Practise answering a stakeholder’s question about the feature’s trade-offs; mentoring and stakeholder communication are explicit requirements.

Your preparation checklist

Prepare a full-stack delivery walkthrough
Choose a project you can discuss end to end. Explain requirements, architecture, your decisions, deployment and outcomes, highlighting relevant backend work in C#/.NET Core, Node.js or Python and a React or Angular frontend with TypeScript.
Show how you use AI assistants across the SDLC
Prepare a concrete example of using GitHub Copilot, Cursor or Claude for tasks such as design, coding, debugging, tests or documentation. Bring examples of how you reviewed outputs and corrected errors; the role emphasizes expert, responsible use.
Sketch an AI-powered feature
As a preparation exercise, design a feature using an OpenAI or Anthropic API. Explain prompt design, where semantic search or an intelligent workflow would fit, and how you would test the feature; do not present the sketch as an NUS project.
Rehearse an application architecture discussion
Draw a simple RESTful application architecture and explain your choices for SQL or NoSQL storage, OAuth2 or JWT authentication, and Docker. Be ready to discuss scalability and security trade-offs relevant to the stated stack.
Collect evidence of engineering quality
Prepare an example showing clean-code decisions, unit, integration and end-to-end testing, CI/CD, and how you identified and resolved a bug or security issue. Distinguish AI-generated suggestions from checks you performed yourself.
Prepare leadership and stakeholder examples
Outline one example of communicating technical decisions to stakeholders and one of mentoring engineers or improving a development practice. Explain what changed and how you measured the result, if you have that evidence.
Write focused questions for NUS IT
Ask which applications and AI features this team would prioritize, how it evaluates AI-assisted code and LLM integrations, and what mentoring would involve. If work arrangements matter, ask for specifics; the posting mentions flexibility but gives no schedule.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…