Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Staff AI Engineer (Fintech): Designing and deploying intelligent, autonomous systems to automate complex financial workflows with an accent on agentic frameworks, orchestration, and pragmatic model selection. Focus on building reliable, scalable production systems that leverage unique financial datasets to eliminate manual operations.
Location: San Francisco (Onsite)
Company
Airwallex is a global payments and financial platform providing integrated solutions for businesses to manage accounts, payments, and treasury at scale.
What you will do
- Architect and deploy AI agents using frameworks like LangChain, LlamaIndex, and AutoGen.
- Apply advanced techniques including prompting, RAG, fine-tuning, and training from scratch.
- Design and optimize agent architectures for reasoning, tool use, and workflow automation.
- Build systems that improve with usage to create compounding AI flywheels.
- Collaborate with backend and frontend teams to integrate AI reasoning into secure production systems.
Requirements
- 5+ years of experience in applied AI/ML engineering with production deployments.
- Strong proficiency in Python and ML frameworks such as PyTorch, JAX, or TensorFlow.
- Proven track record of building and deploying AI agents or multi-step reasoning systems.
- Understanding of monitoring, evaluation, and iteration in production AI environments.
- Ability to balance innovation with reliability, scalability, and compliance.
Culture & Benefits
- Work on high-visibility problems with a global team of over 2,200 innovators.
- Opportunity to build the future of global banking using unique, proprietary financial datasets.
- Collaborative environment that values founder-like energy, ownership, and accelerated learning.
- Commitment to equal opportunity and merit-based hiring.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →