Skip to content
Subject

JamesAbraham

Software Engineering Manager · AI, Data & Platform Engineering

I build the governed plumbing AI products run on, then keep the teams shipping against it.

DomicileDenver, CO
Class codeAI · Data · Platform
Effective2009 — present
Systems on file13 · 9 owned
§1

What I built, wired together

Every box is a system I shipped or own. They are stacked in the order data actually moves through them — intake at the top, the org that ships it at the bottom. Open one to see what it does and what it moved.

Flow keyDataControlAudit evidence
Hover to trace · click to open · arrow keys to walk
1.1OwnedCarrier integrationsCompScience · 2024 — present1.2ShippedVideo ingestion & eventdetectionCompScience · 2024 — 20251.3OwnedAI document extractionCompScience · 2025 — present2.1OwnedMCP data access gatewayCompScience · 2025 — present2.2OwnedAudit trail & SOC 2evidenceCompScience · 2025 — present3.1OwnedAI app platform (Nova)CompScience · 2025 — present3.2OwnedMedallion data platformCompScience · 2024 — present3.3ShippedMonolith modernizationCompScience · 2025 — present4.1OwnedAI in the underwritingworkflowCompScience · 2025 — present4.2ShippedUnderwriting workbenchCompScience · 2024 — 20254.3ShippedClient reportingautomationCompScience · 20255.1OwnedTeam & deliveryCompScience · 2025 — present5.2OwnedAgent toolchain &release pathCompScience · 2025 — present
1Intake
2Governance
3Platform
4Product surface
5Org & delivery
No system openPick a box to read what was built, what it runs on, and what it moved.
§2

Rating worksheet

Numbers I can defend. Each row carries how it was measured, because a figure without a basis is decoration.

Key metrics with the basis of each measurement
ValueMeasureBasisSystem
10xVideo ingestion throughputEvent-driven rearchitecture measured against the legacy pipeline it replaced1.2 ↗
90%Evidence clip processing time cutPer-clip generation time before and after optimization, same input volume1.2 ↗
200+Endpoints moved to typed proceduresUntyped REST surface migrated during monolith modernization3.3 ↗
40%Team efficiency after the pipeline rebuildDelivery throughput after the Linear, LaunchDarkly, Sentry, Trivy and Atlantis pipeline landed5.2 ↗
85%+Sprint completion rate heldRolling average across a 10-engineer distributed team while owning dual P0 initiatives5.1 ↗
15%Revenue lift contributionAttributed to faster risk assessment and renewal workflows in the underwriting workbench4.2 ↗

9 more measures sit on the systems they belong to. Press ⌘K and type a number to find one.

§3

Schedule of capabilities

Architecture and infrastructure first, because that is where I spend the decisions. Marked items are the ones I reach for most; the rest I have shipped with and would again.

Architecture & system designHow the systems are shaped before anyone writes code
Event-driven architectureDistributed systems designAPI designMicroservicesData governance & access controlMonolith decompositionMedallion / lakehouse data architectureHigh availability & fault toleranceAudit and compliance architectureMulti-region PHI/PII data residencyHealthcare interoperability (FHIR)Integration and middleware designSchema and domain modelingTechnical roadmappingArchitecture decision records
Cloud platformsWhere it runs
AWSLambdaECS FargateS3Aurora PostgreSQLRedshiftSQSStep FunctionsDynamoDBRDSEC2CloudFrontIAMAuth0 / OIDCAzure SQLAzure DevOpsDigitalOcean dropletsDigitalOcean Container RegistryDigitalOcean Spaces
Infrastructure & operationsWhat keeps it running
KubernetesDockerTerraformCI/CDObservabilityAnsibleGitHub ActionsAtlantisCaddyGrafanaPrometheusSentryRabbitMQContainer registriesBlue/green and gated releasesInfrastructure as code review
LanguagesWritten in production, not just tried
TypeScriptPythonJavaC#SQLJavaScriptC++PHPBash
Backend & webThe frameworks and protocols in between
Node.jsBuntRPCGraphQLRESTSpring BootHonoExpressMuleSoft / AnypointReact / ReduxAstroNext.jsAngularDrizzle ORMZodWebSockets
AI & MLApplied, in shipped products
LLM application developmentAgentic workflowsModel Context Protocol serversAI data access governanceClaude APIOpenAI APIAI document extractionComputer vision event detectionPrompt and output evaluationRetrieval over governed viewsAgent tooling in CI
DataWarehousing, movement and compliance
Medallion architecture warehousingETL / ELT pipelinesPHI/PII compliance (HIPAA, GDPR)PostgreSQLMySQLRedshiftFHIR and healthcare interoperabilityData governanceAccess classificationPartner and carrier data feeds
DeliveryHow work reaches production
Trunk-based developmentSOC 2 controlsFeature flaggingLinearLaunchDarklySentrySnykTrivyCodeRabbitMixpanelAutomated regressionAudit evidence
LeadershipRunning the team that does it
Engineering managementOnshore and offshore teamsOKR planningVendor and carrier negotiationAgile deliveryMentoringHiring and interviewingCross-team technical directionIncident response

Every item here is searchable. Press ⌘K and type one to see where it was used.

§4

Terms of service

Coverage periods first, most recent at the top. The overlaps are real — the consulting practice ran alongside full-time work for thirteen years. Pick a company to jump to what I did there.

Company

CompScience

AI-driven workers compensation insurance
Software Engineering Manager2024-04 → present · Remote
Promoted 2xSenior Software EngineerLead Senior Software EngineerSoftware Engineering Manager
  • Own the platform AI surface: internal AI app platform, governed agent data access over MCP, document extraction, and AI inside the underwriting workflow.
  • Drove Claude adoption across engineering — most coding and review throughput now runs through it, with agent instructions committed to the repositories.
  • Led the medallion architecture data warehouse ingesting PHI/PII from North American and EU clients under HIPAA and GDPR.
  • Introduced the company audit logging standard behind SOC 2 and made audit instrumentation a condition of shipping.
  • Manage 10 engineers across the US and Vietnam at an 85%+ sprint completion rate and a sub-24-hour review SLA.
  • Before managing, led an offshore team of 8 engineers building the internal tooling that automated client reporting and cut manual overhead 40%.
  • Added GraphQL alongside REST to make the API flexible enough for frontend, backend and data science teams to iterate independently.

Science 37

Decentralized clinical trial platform
Lead Senior Fullstack Engineer2021-05 → 2023-12 · Remote
Promoted 1xSenior Fullstack EngineerLead Senior Fullstack Engineer
  • Promoted to Lead after a year, then led data development for the flagship decentralized clinical trial platform, delivering for partners including Bayer, Johnson & Johnson, and BioMarin.
  • Built middleware and integration layers with Spring Boot microservices and MuleSoft, transforming data into FHIR for EHR consumption.
  • Put EHR data in the hands of clinical teams so remote and in-clinic nurses could run trials, studies, and treatment events on accurate, interoperable patient data.
  • Delivered across React/Redux, Angular, Node.js and Express with CI/CD in Python, Terraform, GitHub, and AWS.
  • Recognized for documentation that sped up issue resolution across teams.

M3 Accounting

Hospitality accounting and business intelligence
Software Developer II2018-03 → 2021-05 · Duluth, GA
  • Designed and built the enterprise data warehouse on Docker, Grafana, Prometheus, RabbitMQ and a C# service, boosting revenue 15% and cutting resource costs, with Ansible automation streamlining operations.
  • Built a secure, standardized enterprise API platform on MuleSoft and Anypoint for partner data exchange, then automated deployments so it scaled with the business intelligence product.
  • Created a multi-format file import service that reduced manual labor costs 20% and integrated cleanly with Marriott, Hilton and IHG Hotels.
  • Delivered the integrations connecting internal applications to external vendor and partner systems, standardizing how data moved between the business and its partners.
  • Earned the MuleSoft Certified Developer credential in 2019, the year MuleSoft CONNECT opened its global roadshow in Atlanta.
  • Helped secure long-term contracts with Trump Towers and Hyatt.

Sage

Enterprise accounting software
Software Developer I2015-11 → 2018-05 · Greater Atlanta Area
  • Resolved software and Pervasive SQL issues for enterprise accounting clients.
  • Supported staging and production deployments and led database query optimizations.

Adult Swim, Turner Broadcasting

Broadcast and streaming media
Software Developer I2013-01 → 2014-10 · Atlanta, GA
  • Built video streaming in JavaScript on a microservices architecture.
  • Shipped mobile games in Unity and web features with React and PHP.

CodePriority LLC

My own company — independent consulting
Founder & Principal Engineer2010-03 → 2023-12 · Independent · Remote
  • Founded and ran my own consultancy alongside full-time engineering roles for thirteen years.
  • Built and maintained 40+ client websites and applications, owning the work end to end — scoping, build, hosting and the client relationship.
  • Consulting engagements with Coinbase on API platform work and with IBM.
  • The brand this site is published under.

IBM

Security services
Cyber Security Analyst2009-01 → 2009-10 · Atlanta, GA
  • Computer forensics, IBM Domino, MS Exchange and Active Directory administration including group policy and permissions.
  • Network security across firewalls, switches, routers, domain administration and topology, on Windows and Linux.
§5

In the open

Most of what I build now is private or internal. What is public is here, pulled from GitHub.

Stars earned278
Followers12
Public repositories1

PhoenixBot

★ 278

Phoenix Bot - A Bird Bot Resurrection

PythonUpdated 2026-07

Everything else

Private

Six more projects run in private repositories, including a live fantasy football platform and two collection apps.

§6

Open a file

I am open to engineering leadership roles working on AI platforms, data infrastructure, or the governance layer between them. Fastest way to reach me is email.