Machine Learning / AI Contractor

KAYGEN•Jersey City, NJ

2d•$60 - $75•Onsite

About The Position

KAYGEN is hiring a Machine Learning / AI Contractor to build and operate production pricing agents and machine-learning solutions. The contractor will own pricing agents on NEO from prototype through production and contribute to the firm's federated portfolio of production agents. The role combines agentic AI and retrieval engineering with pricing and revenue-management models, production MLOps, evaluation, and model-risk documentation. The contractor will build production-ready Graph RAG and vector-search retrieval, design agent memory and organizational context management, develop and operate price-optimization and related models, and compose multi-agent workflows. The role also includes creating agent and model evaluations, maintaining controls and evidence for model-risk review, and deploying solutions on AWS and/or Azure with appropriate security, resiliency, observability, and SDLC practices. The contractor will partner with pricing, product, finance, and business stakeholders across CIB and Payments to deliver supported agents. The position is based in Jersey City, New Jersey, and is listed through April 16, 2027. The stated rate is 60–75; the posting does not specify a currency or rate period.

Requirements

MS in Computer Science, Statistics, Mathematics, Machine Learning, or a related field, or equivalent experience.
Hands-on experience building LLM-powered or agentic applications in production, including tracing, evaluations, and guardrails.
Hands-on experience taking predictive or optimization models to production and maintaining them, including validation, monitoring, and retraining.
Practical RAG experience, including retrieval quality, embeddings, and vector stores.
Strong Python programming skills and deep knowledge of data structures, algorithms, machine learning, data mining, information retrieval, and statistics.
Knowledge of Kubernetes, including AWS EKS.
Experience training models in Databricks and SageMaker.
Experience working with MLflow.
Expert knowledge of at least one of AWS, Azure, or Kubernetes.
Knowledge of data management and data-model design, and real-time processing with SQL databases such as Postgres and NoSQL stores such as OpenSearch and Redis.
Excellent communication skills and ability to work effectively with senior technical and business stakeholders.

Nice To Haves

Graph RAG experience.
Production experience with pricing, revenue management, elasticity, or demand-forecasting models.
Experience with mathematical optimization, including linear and mixed-integer programming (such as OR-Tools or Gurobi), and constrained recommendation problems.
Experience with causal inference, uplift modeling, or experiment design when randomization is limited.
Experience delivering models through model-risk management or an equivalent regulatory review.
Experience in financial services or payments at a large institution, or with commercial card, treasury services, or billing and revenue systems.
Experience with agent frameworks or runtimes, A2A, or MCP.
Experience with agent memory design, including memory nodes and episodic or semantic memory, and organizational context management.
Experience using knowledge graphs and graph databases for retrieval.
Understanding of LLM fine-tuning and small-language-model inference.
Ability to develop full-stack products with modern JavaScript or TypeScript frameworks such as Next.js or Svelte for agent UIs, including AG-UI or NEO UI SDK.
Knowledge of high-performance languages such as Go or Rust.

Responsibilities

Design, build, and ship production pricing agents on NEO, owning them from prototype through production.
Build production-ready Graph RAG and vector-search retrieval over products, rate cards, clients, and pricing policy, including chunking, ranking, and grounding strategies for accurate, auditable explanations.
Design agent memory using episodic and semantic memory nodes, with recall, summarization, and decay policies suited to each use case and safeguards against cross-client information leakage.
Manage organizational context by assembling entitlement-, lineage-, and tenant-aware information so agents access only authorized pricing data.
Build and productionize ML training pipelines for price optimization, elasticity, win probability, and deal scoring, including feature engineering over billing, transaction, and deal-history data on Databricks.
Build batch and online inference for models, meeting seller-conversation latency needs and ensuring scored recommendations are reproducible.
Validate models against realized prices; develop challenger models; measure uplift and elasticity; and monitor production for drift, margin erosion, and revenue leakage.
Compose multi-agent workflows using A2A and integrate tools and data through MCP servers, including pricing, CRM, billing, data, infrastructure, and collaboration systems.
Build task-level and end-to-end agent evaluations, regression suites, LLM-as-judge assessments, numerical-correctness and explanation-faithfulness checks, and release quality and safety gates.
Produce and maintain model documentation, test evidence, and controls required for model-risk review, including as models are retrained.
Deploy and operate solutions on public cloud using strong SDLC, security, resiliency, and observability practices.
Partner with pricing, product, finance, and business stakeholders across CIB and Payments to turn use cases into shipped, supported agents.