Machine Learning Engineer

Location: San Francisco, California – in-person, approximately six days per week (Sunday to Friday, with some weekend flexibility)

Employment: Full-time

Visa sponsorship: O-1 and OPT

Salary: $140,000–$300,000 + competitive equity

About the company

We are partnering with an ambitious, early-stage fintech startup building the financial infrastructure for the agentic web. As AI agents increasingly interact with businesses, make purchasing decisions and execute transactions on behalf of users, there is a growing need for secure and reliable payment infrastructure.

We are looking for a Founding AI/ML Engineer specialising in Authorization to take ownership of the intelligence behind how AI agents make purchases.

Imagine a user telling an AI agent:

You'll design and build the models that determine whether an agent's actions are authorised, how risky they are, and when additional user approval is required.

Starting as a hands-on individual contributor, you'll own the modelling strategy and have the opportunity to build and lead the AI/ML function as the company scales.

What you'll be doing

• Own model development: Design and build the data pipelines and modelling architecture behind the company's two core models, including a small language model focused on product and shopping-cart context.
• Build authorisation intelligence: Translate natural-language user approvals into structured, enforceable authorisation rules covering merchants, spending limits, products, substitutions and time constraints.
• Detect suspicious behaviour: Develop models to identify deviations from user intent, unauthorised substitutions, prompt injection attempts and excessive spending before transactions are executed.
• Develop risk-scoring systems: Combine agent identity, user history, device information and merchant signals to determine whether transactions should be approved, challenged or rejected.
• Own model evaluation: Build evaluation frameworks, data labelling processes and performance metrics, with particular focus on false approvals and false declines.
• Deploy at scale: Ship production-ready models directly into the payment authorisation flow, optimising for accuracy, reliability, inference latency and cost.

What we're looking for

• Strong experience in Machine Learning, Applied AI or AI/ML Engineering.
• Hands-on experience developing and deploying production ML models.
• Experience with LLMs, Small Language Models (SLMs), NLP or agentic AI systems.
• Strong Python and experience with modern ML frameworks such as PyTorch, TensorFlow or Hugging Face.
• Experience with anomaly detection, fraud detection, risk modelling, trust and safety or decision-making systems.
• Understanding of model evaluation, experimentation, data pipelines and performance optimisation.
• Experience deploying low-latency models and production inference services.
• Ability to take ownership of technical problems, make architectural decisions and build systems from the ground up.

Particularly interesting backgrounds include: Fintech, payments, fraud detection, transaction authorisation, AI agent security, LLM evaluation and real-time risk modelling.

Why join?

• Founding-level ownership: Take responsibility for a critical part of the company's AI/ML architecture from day one.
• Cutting-edge problem: Help define how autonomous AI agents interact with financial systems and make purchases safely.
• Technical influence: Shape the modelling strategy and build the foundations of the AI/ML team.
• Competitive package: $140K–$300K base salary plus competitive equity.
• High-impact environment: Join a small founding team preparing to launch a new financial infrastructure platform.