Principal AI/ML Research Engineer

WEX•California, MD

2d•$250,300 - $289,000•Hybrid

About The Position

As a Principal AI/ML Research Engineer, this technical leader will drive applied AI research, novel model development, and algorithmic innovation across generative AI (GenAI), deep learning, and traditional machine learning. This technologist will lead the discovery, design, and prototyping of state-of-the-art architectures to solve complex, high-scale commerce and fintech challenges. For example, this work may include pioneering Transformer-based Payment Foundation Models (PFMs)—adapting self-attention mechanisms to large-scale tabular transaction data, user behavioral sequences, and multi-modal financial streams to produce universal embeddings that power a wide array of downstream commerce applications. This role bridges state-of-the-art academic/industry research with real-world production impact. You will closely partner with leaders in Data Engineering, Product, Security, Risk & Compliance, and Line of Business (LOB) technical teams to identify high-value opportunities, benchmark novel architectures (e.g., Tabular Transformers, LLMs, RAG, Reinforcement Learning), and transition experimental models into viable production pipelines. This Principal AI/ML Research Engineer will hold technical ownership of WEX’s AI research strategy, model optimization, algorithmic rigor, and AI experimentation standards. The vision behind WEX’s AI research is to transform multi-modal commerce and payment data into intelligent, predictive models that drive competitive advantage. This role reports to the VP of Data Lake and AI Engineering located in the San Jose, CA Bay Area, but can be located in Seattle, WA; Portland, ME; Boston, MA; or Chicago, IL. The ideal candidate is a hands-on technical leader with deep domain knowledge in applied AI/ML research, statistical modeling, and experimental design, combined with strong strategic vision and communication skills.

Requirements

12+ years of experience in software/ML engineering, with 5+ years dedicated to applied AI/ML research, model architecture design, novel algorithm development at scale, AI application development, and AI agent development.
Master’s or Ph.D. in Computer Science, Machine Learning, Statistics, Mathematics, or a related quantitative field preferred, or equivalent qualifications.
Deep expertise applying Transformer models, self-attention mechanisms, and self-supervised pre-training techniques to sequential, time-series, or tabular financial/transactional datasets.
Proven expertise in modern AI paradigms—Transformers, LLM pre-training/fine-tuning (LoRA, PEFT), RAG architectures, prompt engineering, Diffusion, or Reinforcement Learning (RL/RLHF).
Strong theoretical foundation and hands-on experience in supervised/unsupervised learning, time-series forecasting, anomaly detection, and graph algorithms.
Expert proficiency in Python and deep learning frameworks (PyTorch, TensorFlow, JAX). Experience in C++ or Java for performance-critical ML components is a plus.
Strong understanding of distributed training/inference frameworks such as Ray, DeepSpeed, Megatron, or Spark.
Hands-on experience with cloud environments (AWS/Azure) and ML platforms like SageMaker, MLflow, Databricks, or vector databases (LanceDB, Pinecone, Qdrant, Milvus).
Demonstrated ability to translate complex academic literature into production-grade prototypes. Publications in top AI/ML venues (NeurIPS, ICML, KDD, ACL, etc.) or open-source contributions are highly desirable.
Experience applying AI/ML algorithms to payments, fintech, risk management, fraud detection, or transactional big data is a major plus.
Exceptional capability to explain highly technical research concepts and mathematical models to non-technical executive leadership and cross-functional partners.
Possesses a relentless drive to stay ahead of the AI curve, continuously learning and experimenting with cutting-edge techniques.
Bridges the gap between research curiosity and business execution with humility, empathy, and transparent communication.
Thrives in a fast-paced environment, comfortably pushing boundaries, challenging the status quo, and driving adoption of novel methods through influence and partnership.
Demonstrates an uncompromising commitment to AI fairness, safety, explainability, and regulatory compliance.

Nice To Haves

Experience in C++ or Java for performance-critical ML components is a plus.
Publications in top AI/ML venues (NeurIPS, ICML, KDD, ACL, etc.) or open-source contributions are highly desirable.
Experience applying AI/ML algorithms to payments, fintech, risk management, fraud detection, or transactional big data is a major plus.

Responsibilities

Design self-supervised pre-training strategies (e.g., masked transaction prediction) on raw payment histories to generate multi-purpose user/entity embeddings, enabling efficient fine-tuning for fraud detection, credit risk, dispute prediction, and authorization optimization.
Drive applied research in adapting Transformer architectures (encoder/decoder, self-attention mechanisms, and Tabular Transformers) to build enterprise-grade Payment Foundation Models (PFMs) tailored to transaction streams.
Lead applied research in AI/ML to solve high-impact business problems in fraud detection, risk scoring, predictive commerce, customer engagement, and automated decision-making.
Spearhead research and prototyping in modern AI frameworks—including LLM fine-tuning, retrieval-augmented generation (RAG), agentic workflows, multi-modal systems, and Reinforcement Learning.
Lead development of effective framework and methodologies in auto AI agent feedback collection and agent self-learning and self-improvement.
Lead AI model optimization for performance, latency, and cost for Wex use cases, including developing model routers.
Lead the development of effective AI agent evaluation methodologies.
Design and execute rigorous rapid-prototyping pipelines to validate new model architectures, feature representations, and algorithms before handing off to ML Engineering for scale-out.
Serve as a subject matter expert in state-of-the-art AI techniques, publishing research internally/externally where applicable, monitoring the academic landscape, and identifying emerging technologies to keep WEX at the forefront of AI innovation.
Partner closely with Data Science, ML Engineering, Risk, Security, and Product teams to align research agendas with long-term business goals and ensure responsible AI deployment.
Establish baseline benchmarks, mathematical validation protocols, explainability (XAI) frameworks, and performance evaluation metrics across predictive accuracy, latency, fairness, and bias reduction.
Collaborate with Information Security, Compliance, and Governance teams to ensure AI models comply with data privacy regulations, ethical AI principles, and robust security standards.
Set a high standard for scientific research and engineering rigor within the team. Provide technical guidance, code/math reviews, and mentorship to engineers and data scientists across the organization.
Define, prioritize, and execute WEX’s AI research roadmap, balancing foundational research with near-term business impact and clear OKRs.