# Oxford Global - Senior GenAI / Agentic Engineer

## Source

- Original JD posting: https://files.keithsteward.com/Oxford_Global-Senior_GenAI_Agentic_Engineer.pdf
- Source fetched and normalized: 2026-08-18 9:18 AM EDT
- Source PDF SHA-256: `1255382079ca8c94c1814736f5cb755a56e27b720de9341a63d3d5281f304d78`
- Posting time: not disclosed in the source PDF
- Applicants: not disclosed

## Normalized Posting Metadata

- Recruiter/staffing company: Oxford Global
- Client: undisclosed healthcare and health-insurance organization
- Role: Senior GenAI / Agentic Engineer
- Employment type: contract; exact classification not disclosed
- Start: September 1, 2026
- Duration: through the end of 2026, with extension likely through 2027
- Work model: primarily remote, Eastern Time business hours
- Travel: potential for one or two trips to New Jersey; trip duration not disclosed
- Compensation: not disclosed
- Team: one or two engineers joining a small technical pod

## Overview

Oxford Global seeks a Senior GenAI / Agentic Engineer to help a healthcare and health-insurance client scale an AI Marketplace across 31 organizations. The marketplace will allow participating organizations to discover, test, share, and use AI-powered services. It will include a sandbox where organizations can bring their own data, test solutions, and evaluate use cases. Following successful pilots, the client is scaling the platform and adding one or two engineers to a small technical pod.

## Responsibilities

- Design and build production-grade GenAI and agentic AI solutions for healthcare-payer workflows.
- Develop AI agents and workflows using Amazon Bedrock, AgentCore, and the broader AWS AI/ML stack.
- Scale successful pilots into reusable, enterprise-grade AI capabilities.
- Support marketplace and sandbox functions where members test solutions with their own data.
- Translate healthcare workflows into scalable AI and agentic solutions.
- Contribute to architecture, technical design, development, testing, implementation, and engineering best practices.
- Collaborate with AWS Professional Services, Accenture, BCBSA, and other technical teams.

## Key Use Cases

- Prior authorization
- Claims adjudication and processing
- Automated remediation
- Healthcare payments
- Member and customer experience

## Required Qualifications

- Senior-level experience developing Generative AI, LLM, or agentic AI applications.
- Hands-on experience building AI agents and agentic workflows.
- Strong AWS AI/ML experience, particularly Amazon Bedrock and AgentCore.
- Experience with SageMaker and related AWS AI/ML technologies.
- Strong software-engineering and solution-development skills.
- Experience taking AI solutions from pilot or prototype through production.
- Ability to work hands-on across architecture, development, and implementation.

## Preferred Qualifications

- Healthcare-payer experience is strongly preferred.
- Experience with claims, prior authorization, adjudication, or healthcare payments.
- Experience building AI solutions in regulated environments.

## Evidence Map

- Production GenAI applications — direct, recent evidence: Keith architected and helped build AssistX, a production GenAI platform with 70+ automations across more than 20 workflows, released in eight months.
- Agentic workflows — direct, recent evidence: AssistX uses AI agents, multi-agent architectures, RAG, LangChain, LangGraph, workflow orchestration, and human-controlled safeguards.
- AWS AI/ML — direct evidence: four years in AWS specialist roles, advising 200+ enterprise customers and working with SageMaker, TensorFlow, EMR, Glue, Lambda, S3, and other AWS services.
- Bedrock and AgentCore — source-supported skills, but the career sources do not provide a quantified production outcome specifically tied to AgentCore; scored as direct but less measurable.
- Software engineering and architecture — direct evidence across full-stack AssistX development, AWS architectures, cloud platforms, APIs, Python, FastAPI, PostgreSQL, and production operations.
- Pilot-to-production delivery — direct evidence from AssistX, SuccessKPI, MassMutual, TriMark, NorthBay, and earlier commercial software products.
- Regulated environments — direct evidence from MassMutual privacy and AI-governance work, HIPAA-regulated dental-health software, and FDA-cleared health technology.
- Healthcare-payer workflows — adjacent evidence only. Keith has supported insurance/financial-services and healthcare contexts, but the sources do not establish direct prior-authorization, claims-adjudication, or healthcare-payment operating experience.

## Fit Gate

- Outcome: BORDERLINE
- Overall weighted fit: 78.6%
- Capability coverage before compensation/title gating: approximately 87%
- Positioning track: Technical IC
- Central mandate: hands-on development and production scaling of AWS-based GenAI and agentic systems.
- Strongest differentiators: current production agentic-AI platform development; deep AWS AI/ML and SageMaker background; repeated prototype-to-production delivery.
- Material gaps: compensation is undisclosed for a title that is not Senior Director, VP, or Chief; no direct healthcare-payer workflow evidence; AgentCore depth is source-supported but not outcome-backed.
- Overqualification risk: material but potentially manageable because Keith remains hands-on in architecture and development; executive branding must not appear in the top third of an IC resume.
- Decision: Keith explicitly overrode the BORDERLINE gate on August 18, 2026 and requested generation of the resume and cover letter.

## Industry-Claim Integrity

- Target industry: healthcare payer / health insurance.
- Any future materials must not claim direct healthcare-payer, prior-authorization, claims-adjudication, or healthcare-payment experience.
- Source-supported regulated-healthcare, insurance/financial-services, privacy, AI-governance, and production-engineering evidence may be presented accurately without converting it into payer-domain experience.

## Keyword Diagnostic

- 14 of 18 priority concepts are directly supported or credibly transferable: 77.8%.
- Unsupported or only adjacent concepts: healthcare-payer operations, prior authorization, claims adjudication, and healthcare payments.

## Artifact Metadata

- Archived JD capture generated: 2026-08-18 9:18 AM EDT
- Application materials generated after Keith's explicit BORDERLINE override: 2026-08-18 9:33 AM EDT
- Resume: `Keith_Steward_Senior_GenAI_Agentic_Engineer_resume.pdf`
- Cover letter: `Keith_Steward_Senior_GenAI_Agentic_Engineer_cover_letter.pdf`
- Resume format: two-page, single-column, ATS-readable Letter PDF
- Cover-letter format: one-page Letter PDF
- Validation: content, industry-claim integrity, active hyperlinks, page geometry, bullet geometry, text extraction, and visual layout passed
- Google Drive: not used