Machine Learning Engineer

About Payle
Payle is building the financial infrastructure layer for the AI agent economy. Our platform enables AI agents to transact under human-defined rules: programmable spending policies, deterministic authorization, and a hash-chained append-only ledger that makes every transaction verifiable.
We are a YC applicant backed by an independently audited, production-grade authorization engine. Our engineering culture is built on three non-negotiable principles: deterministic code decides where money moves, every decision is traceable to its source, and nothing ships without tests that prove it.
We operate as a remote-first, async-first distributed team across four countries. We hire by artifact: every engineer and researcher ships a working challenge before we discuss terms.

The Role
We are seeking a Senior Machine Learning Engineer to design, build, and own the risk and trust scoring systems that underpin autonomous payment authorization. This is a foundational role: you will define how Payle evaluates risk for machine-initiated financial transactions, establish the ML infrastructure from first principles, and set the standards for model governance, explainability, and fairness that the company operates by going forward.
You will report directly to the CTO and work closely with the CEO on product strategy, regulatory alignment, and the technical roadmap for the risk and credit domains.

What You Will Build
Risk Scoring Systems

1. Design and deploy real-time transaction risk scoring for AI-agent-initiated payments, with a hard latency budget of 20ms at p99
2. Build fraud detection models covering velocity anomalies, amount outliers, new-merchant risk, structuring patterns, and behavioral deviations from agent baselines
3. Develop agent trust scoring: longitudinal models that evaluate an agent's historical transaction behavior, outcome verification rate, policy adherence, and dispute history

Feature Engineering Infrastructure

1. Design and implement an offline/online feature store with point-in-time correctness guarantees, backed by our append-only transaction ledger
2. Build feature pipelines from raw event streams (authorization, outcomes, chargebacks, policy violations, merchant interactions) with automated freshness and consistency validation
3. Establish feature lineage, documentation standards, and automated leakage detection integrated into CI

Model Development and Governance

1. Train, evaluate, and deploy models starting with gradient-boosted trees (LightGBM/XGBoost) on tabular transaction data, evolving toward more complex architectures as labeled data volume supports
2. Implement SHAP-based explainability for every production decision: per-transaction factor attribution with stable, auditable output
3. Establish model versioning, registry, and staged deployment (shadow mode → canary → full production) with documented rollback procedures
4. Build drift monitoring: population stability index (PSI) tracking per feature, automated alerting on distribution shift, and scheduled retraining pipelines triggered by drift thresholds

Credit and Underwriting Models (Future)

1. Design behavioral underwriting models for agent-initiated credit (FundingFlex): probability of default, exposure at default, and loss given default models trained on agent transaction history
2. Build adverse action reason code generation: every credit decision must produce machine-generated, human-readable justification codes compliant with regulatory requirements
3. Implement disparate impact testing and fairness metrics (demographic parity, equalized odds) on all underwriting-adjacent models, with documented thresholds and review cadence

What We Look For
Required:

1. 5+ years building and deploying ML systems in production, with at least 2 years on fraud detection, risk scoring, credit underwriting, or a related financial services domain
2. Strong Python with production ML frameworks: scikit-learn, LightGBM/XGBoost, pandas, NumPy
3. Deep understanding of point-in-time feature correctness, target leakage, and the failure modes of offline-online feature skew
4. Experience with model explainability (SHAP, LIME, or equivalent) and the ability to translate model outputs into business-actionable insights
5. SQL proficiency for feature extraction from transactional databases at scale
6. Experience with model versioning, A/B testing infrastructure, and staged rollout strategies
7. Master's or PhD in a quantitative field (Computer Science, Statistics, Applied Mathematics, Physics, Economics) or equivalent demonstrated experience

Preferred:

1. Experience with on-device or edge ML deployment
2. Familiarity with GDPR Article 22 (automated decision-making), ECOA/Reg B adverse action requirements, or EU AI Act risk classification
3. Experience building ML systems for payment networks, digital wallets, or BNPL platforms
4. Publications or open-source contributions in fraud detection, anomaly detection, or credit risk modeling

What We Offer

• Compensation: $60,000-90,000/year equivalent, structured as monthly payments (part-time to start, full-time at YC acceptance)
• Equity: 0.5-1.0% (4-year vesting, 1-year cliff), granted upon full-time conversion
• Ownership: You define the ML infrastructure, the model governance framework, and the evaluation standards. You are the first and senior ML voice at Payle.
• Data: A proprietary behavioral dataset every AI agent transaction with full context, outcome, and policy trace that no credit bureau, payment network, or fraud consortium has ever collected
• Impact: Your models directly gate financial transactions. Your explainability framework sets the standard for regulatory compliance. Your fairness testing establishes the company's ML ethics baseline.
• Team: A distributed, artifact-driven engineering culture. Direct access to the CEO and CTO. No layers, no bureaucracy, no meetings that could have been a Slack message.