Senior Software Engineer, Machine Learning Platform

AIRWALLEX (SINGAPORE) PTE. LTD.; Singapore; On-site; SGD 180,000 - 240,000; Full time

Job summary

Airwallex (Singapore) PTE. LTD. is seeking an experienced software engineer to build and scale the machine learning platform that powers risk decisioning across Airwallex.

You will design data and model infrastructure, productionise ML workflows, and improve experimentation speed and decision quality. You will partner with ML engineers, data scientists, PMs and risk specialists to turn complex fraud problems into dependable systems, and you’ll be based in Singapore, working from the office five

Qualifications

• 5+ years in software engineering with focus on model training infra.
• Bachelor’s or Master’s in CS, Engineering or a related field.
• Hands-on with PyTorch/TF/JAX and model training engines.

Responsibilities

• Build and scale the ML platform powering risk decisioning across Airwallex.
• Design reliable data and model infrastructure; productionise ML workflows.
• Collaborate with ML engineers, data scientists, PMs and risk experts to tackle fraud problems.
• Based in Singapore; work from the office five days a week.

Skills

Model training infra

Deep learning

Python

Distributed systems

GPU hardware

Education

CS/Engineering degree

Tools

Kubernetes

Kubeflow

Ray

Airflow

Job description

How you'll make impact

You'll build and scale the machine learning platform that powers risk decisioning across Airwallex, helping teams develop, deploy, monitor, and improve models that protect every dollar moving through our platform.

You'll design reliable data and model infrastructure, productionise machine learning workflows, and improve the speed and quality of experimentation and decisioning across the Risk Platform.

You'll partner closely with machine learning engineers, data scientists, product managers, and risk specialists to turn complex fraud and risk problems into dependable systems.

You'll be based in Singapore and work from the office five days a week.

What we're looking for

Essentials

• 5+ years of software engineering experience, with at least 3+ years focused on model training infrastructure, model serving systems, or MLOps platforms.
• Bachelor's or Master's degree in Computer Science, Engineering, or a related technical field.
• Hands-on experience with modern deep learning frameworks (e.g., PyTorch, TensorFlow, JAX) and model training execution engines.
• Strong proficiency in core programming languages such as Python, Java, or C++.
• Experience with distributed orchestration and workflow management tools (e.g., Kubernetes, Ray, Kubeflow Pipelines, Airflow).
• Solid understanding of GPUs, including GPU architecture, hardware acceleration, and GPU-based training or inference optimization.

Preferred

• Experience with model acceleration frameworks and Large Language Models (LLMs).
• Proficiency in performance profiling and bottleneck identification using tools like NVIDIA Nsight Systems for training and inference optimization.
• Experience with cloud platforms (e.g., AWS, GCP) and building large-scale, low-latency production machine learning infrastructure.

You'll thrive here if

• You're comfortable owning the roadmap yourself.
• You own the outcome and you don't wait for permission to fix what's broken.
• You're comfortable with ambiguity. Give you a problem, not a prescription, and you'll run with it.
• You enjoy working closely with people across multiple countries and time zones as part of one connected, global team, including flexing your hours occasionally to make that connection work.
• You value in-person collaboration and are happy being in the office five days a week.

Learn more about your team

Risk Platform builds the decisioning infrastructure that sits between Airwallex and every dollar that moves through it, protecting 150,000+ businesses moving over US$260 billion a year across 200+ countries and 90+ currencies, and deciding, often in milliseconds, whether a new signup is real, a payment is safe, or a login is who they claim to be. The hard part is that fraud evolves fast, and every decision carries a two-sided cost: miss an attack and money is lost, over-block and a legitimate business cannot get paid. We build this with streaming pipelines processing billions of events a day, graph databases exposing coordinated fraud rings, ML models scoring every transaction, and LLM agents that triage alerts. You don't need a fintech background, just an appetite for adversarial systems problems where the scoreboard is measured in dollars. If you want to help scale one of the world's fastest-growing financial platforms safely, this is the team.