AI R&D Engineering Intern — AI Infrastructure, Model Routing & Applied Research

A-Field Tech; Hong Kong; Hybrid; HKD 89,000 - 134,000; Full time

Job summary

NexLLM, a Hong Kong-based AI infrastructure company, seeks a University student for a full-time AI R&D Engineering Intern role. You will work with the founder and engineering team on real AI infrastructure projects, prototyping, evaluation, and product improvements for a live platform with multiple model providers.

Learn model routing, evaluation, and API integrations while contributing to production-ready work.

Qualifications

• Current university enrolment in computer science, software engineering, AI, ML, data science, math, or related field.
• Strong interest in AI models, software engineering, or developer infrastructure.
• Basic programming experience in Python, TypeScript, JavaScript, Go, Java, C++, or a similar language.
• Familiarity with APIs, Git, databases, or cloud services.
• Ability to break unclear problems into smaller steps.
• Willingness to test assumptions and measure results.

Responsibilities

• Researching and prototyping model-routing strategies.
• Comparing models across quality, cost, latency, availability, and reliability.
• Building benchmark datasets and reproducible evaluation tools.
• Developing fallback, retry, timeout, caching, and quota-management features.
• Investigating provider errors, rate limits, response formats, and performance changes.
• Integrating new model providers and standardising their APIs.
• Building scripts, dashboards, data pipelines, and internal developer tools.
• Turning validated experiments into well-tested product features.

Skills

University student

Python

TypeScript

Go

JavaScript

Java

C++

APIs

Git

Databases

Cloud services

Education

Current university enrolment in CS/SE/AI

Tools

Git

Databases

Cloud services

APIs

Job description

Help build the intelligence layer for Asia's AI infrastructure

A-Field Tech Limited is a Hong Kong based AI and network infrastructure company. NexLLM is a Hong Kong based AI infrastructure company building a stable and secure regional gateway for developers who need reliable access to leading AI models.

Our live platform provides developers with OpenAI-compatible APIs across a growing catalogue of models, including Claude, OpenAI, Gemini, Qwen, and DeepSeek. We are now developing the next layer: intelligent model routing, model evaluation, reliability tooling, and developer infrastructure that helps applications choose and operate the right model for each task.

We are looking for a full-time AI R&D Engineering Intern to join our engineering team. This internship is designed for a university student who wants practical experience working on real AI infrastructure, applied research, and software engineering problems.

You will work closely with the founder and engineering team on projects that affect a live platform. This is not a purely observational internship or a role limited to administrative tasks. You will be expected to learn quickly, build working prototypes, measure results, and contribute to production improvements with appropriate guidance.

Internship details

• Role: AI R&D Engineering Intern
• Commitment: Full-time internship
• Location: Hong Kong / flexible hybrid arrangement
• Company: NexLLM
• Team: AI infrastructure, model routing, evaluation, and platform engineering
• Duration: To be agreed based on the university programme and candidate’s availability
• Start date: Flexible, subject to mutual agreement

The work may involve Python, TypeScript, JavaScript, Go, HTTP APIs, databases, asynchronous systems, cloud services, data analysis, and AI model APIs. You will not be expected to know everything before joining; we care about your ability to learn and make steady, evidence-based progress.

What You May Work On

• Researching and prototyping model-routing strategies
• Comparing models across quality, cost, latency, availability, and reliability
• Building benchmark datasets and reproducible evaluation tools
• Developing fallback, retry, timeout, caching, and quota-management features
• Investigating provider errors, rate limits, response formats, and performance changes
• Integrating new model providers and standardising their APIs and capabilities
• Building scripts, dashboards, data pipelines, and internal developer tools
• Reviewing relevant papers, technical documentation, and open-source projects
• Turning validated experiments into well-tested product features
• Writing clear technical notes explaining methods, results, limitations, and recommendations

What You Will Learn

• Building and debugging software used by real developers
• Working with multiple commercial and open-source AI model providers
• Designing experiments and evaluating model behaviour
• Measuring latency, cost, quality, failure rates, and recovery performance
• Building reliable API integrations and distributed services
• Understanding how AI gateways, routing systems, and token usage work
• Translating research ideas into practical engineering decisions
• Communicating technical findings clearly
• Working in an early-stage company where priorities evolve quickly

How The Internship Will Work

• Understand a technical problem and the user or product context.
• Review existing approaches, documentation, and relevant research.
• Form a hypothesis and define how success will be measured.
• Build a small experiment, prototype, or test harness.
• Analyse the results and identify limitations.
• Discuss the findings with the engineering team.
• Improve the solution and, where appropriate, help move it into production.
• Document what was learned and what should happen next.

What We Are Looking For

We are looking for a university student who is curious, technically capable, and motivated to build real systems.

You Should Have

• Current university enrolment in computer science, software engineering, artificial intelligence, machine learning, data science, mathematics, or a related field
• Strong interest in AI models, software engineering, or developer infrastructure
• Basic programming experience in Python, TypeScript, JavaScript, Go, Java, C++, or a similar language
• Familiarity with APIs, Git, databases, or cloud services, through coursework or personal projects
• Ability to break an unclear problem into smaller steps
• Willingness to test assumptions and measure results rather than rely on guesses
• Care with details, reproducibility, and technical documentation
• Ability to communicate clearly and work collaboratively
• Availability for a full-time internship during the agreed internship period

Prior professional experience is not required. Strong coursework, personal projects, research projects, hackathon work, open-source contributions, GitHub repositories, or technical writing are welcome.

You do not need prior experience with AI gateways, large language model infrastructure, or distributed systems. We value learning ability and practical curiosity more than a long list of tools.

What Makes Someone a Strong Fit

• Like understanding how things work rather than only using them
• Enjoy both reading and building
• Can turn a question into a small experiment
• Are willing to start with an imperfect prototype and improve it
• Care about whether your results are reliable and reproducible
• Take responsibility for finishing work and communicating blockers early
• Want to see your work used in a real product
• Are excited by the fast-moving AI infrastructure landscape in Asia

This internship may not be the right fit if you are looking only for a passive learning experience, do not want to write or debug software, or prefer a fixed task list with no ambiguity. Early-stage engineering work requires curiosity, initiative, and a willingness to learn unfamiliar systems.

Why join NexLLM

• Work on a live AI platform used by developers
• Gain practical experience with multiple leading AI models and providers
• Work directly with the founder and engineering team
• Learn about model routing, evaluation, reliability, APIs, and cloud infrastructure
• Build projects that can become real product capabilities
• Receive feedback and guidance from experienced team members
• Explore the intersection of applied AI research and product engineering
• Help build a regional AI infrastructure capability from Hong Kong
• Potentially be considered for future engineering opportunities based on performance and business needs

Compensation And University Coordination

This is a full-time internship. Internship compensation, working arrangement, duration, start date, and any university documentation will be discussed with shortlisted candidates and confirmed before the internship begins.

Candidates who need an internship agreement, supervisor information, learning objectives, progress reviews, or other university documentation should mention this when applying.