Graham Anderson
Evergreen, CO · gramman87@gmail.com · grahamanderson.dev · linkedin.com/in/graham-anderson-denver
Based in Evergreen, CO · open to relocation
Summary
Full-stack software engineer who builds production features across the whole stack, JVM/Spring Boot services, APIs, and data models on the backend, TypeScript and React on the front end. Ships agentic AI applications on Claude in Python and TypeScript: MCP servers (producer and consumer), tool-calling agents, sub-agent orchestration, RAG, and real-time streaming. Modernizes legacy enterprise systems and operates production services on AWS, OpenShift, and Kubernetes, owning them through deployment, incident response, and reliability. Takes pride in well-tested, maintainable code and sound API governance: versioning, backward compatibility, and security on public-facing surfaces.
Core Strengths
- ·Backend (JVM): Java, Spring Boot, REST APIs, microservices, OpenAI-compatible endpoints, systems integration, SQL
- ·Frontend: TypeScript, React, Next.js, HTML/CSS, Tailwind, data visualization
- ·AI & Agentic: Claude API, tool & function calling, MCP (producer + consumer), agents and sub-agents, real-time streaming (SSE/WebSockets), RAG, evaluation harnesses
- ·Testing & Quality: JUnit, Claude-as-judge evaluations, evaluation harnesses; API governance: versioning, backward compatibility, security standards
- ·Cloud & DevOps: AWS, Kubernetes, OpenShift (OCP), Docker, GitLab CI/CD, HashiCorp Vault, full service-lifecycle ownership and incident response
- ·Delivery: Python, Agile/Scrum, cross-functional collaboration, release planning, $80M+ program leadership
Professional Experience
- ·Modernized a legacy federal application by containerizing it and migrating it onto OpenShift/Kubernetes, implementing Java/Spring Boot microservices and owning the services through deployment, incident response, and ongoing reliability work.
- ·Build and consume REST APIs across a microservice architecture, upholding versioning, backward compatibility, and security standards on public-facing surfaces.
- ·Enhanced GitLab CI/CD pipelines through parallelization and caching, reducing deployment time by 40% and tightening the loop between a code change and a deployable build.
- ·Implemented secure secrets management with HashiCorp Vault, strengthening security posture and meeting federal compliance requirements.
- ·Partner across product, architecture, and client teams to translate ambiguous requirements into well-scoped, tested engineering work.
- ·Developed Java-based applications with SmartGWT/JavaScript front-ends for deployment on embedded/IoT devices serving enterprise data-center customers.
- ·Authored and consumed RESTful APIs supporting real-time device communication and integration with adjacent enterprise systems.
- ·Streamlined deployment workflows by optimizing integration scripts, improving release efficiency and reducing manual handoffs.
- ·Built full-stack applications in Java, Spring Boot, and JavaScript deployed on AWS with RESTful service architectures.
- ·Served as Scrum Master and Database Administrator, enforcing Agile cadence, facilitating ceremonies, and driving robust schema design.
- ·Led pre-construction on $80M+ commercial and industrial programs, owning scope development, estimating, business-case development, procurement strategy, and risk evaluation before mobilization.
- ·Evaluated effort, risk, and priority across competing workstreams to build delivery roadmaps, the same trade-off calls that drive release planning on an engineering team.
- ·Coordinated across procurement, engineering, manpower, and scheduling functions, building the cross-functional collaboration muscle that full-stack delivery demands.
Agentic AI Engineering: Independent Work
- ·Spring Boot MCP Agent (github.com/Gramman87/spring-mcp-agent): a full-stack Java 21 / Spring Boot service that publishes tools over an MCP producer surface and runs a streaming agent consuming them over SSE, fronted by a React/TypeScript UI, with an OpenAI-compatible endpoint and a JUnit suite covering tools, registry, agent logic, and controllers. The LLM seam is isolated so a real Claude tool-use call drops straight in.
- ·Ship full-stack applications on the Claude API: MCP servers as both producer and consumer (stdio + Streamable HTTP transports), tool-calling agents, sub-agent orchestration, RAG pipelines, and real-time streaming UIs in Python and TypeScript/React. Every project is live and open-source.
- ·Stand up evaluation harnesses, including Claude-as-judge scoring on routing correctness, coverage, and output quality across hand-written cases, to iterate prompts and tool definitions and catch regressions.
- ·Hands-on with agentic coding tools (Claude Code, Cursor) and LLM integration patterns: tool/function calling, OpenAI-compatible endpoints, data streaming, and the Model Context Protocol as a producer and consumer.
Certifications
- ·Kubernetes: Cloud Native Ecosystem
- ·Spring Boot 1.0 Essential Training
- ·Software Architecture Foundations
Education
- ·Skill Distillery: Certificate, Java Full Stack Development (2021–2022)
- ·Emily Griffith Technical College: Certificate, Electrical Apprenticeship (2008–2012)
- ·Metropolitan State University of Denver: Computer Science (coursework)