AI Engineer · Full-Stack Developer

I build AI systems
that ship.

Two multi-agent LLM platforms in production right now. Real users, real billing, real infrastructure. Looking for the team worth joining.

2,600+ Commits
2 Apps in Production
34 Edge Functions
6,000+ RLHF Evaluations
01 / The Arc

Four years from evaluation
to production.

2021-23
Multimodal production at scale. 5,000+ images in Midjourney. 984 AI-generated songs distributed across major platforms. Seasonal art packages at 8K resolution for Samsung Frame TV. Learned what compounds and what breaks when you push generative systems past demo stage.
2024
Started evaluating LLM outputs at scale. 6,000+ RLHF evaluations across Claude, GPT, and Gemini model families. Systematic red teaming for hallucinations, bias vectors, jailbreak vulnerabilities. 2,000 hours of hands-on interaction. Built the intuition that informs everything I ship now.
2025
Started shipping production software. Two SaaS platforms from zero. 47 development phases. A billing system rewritten three times before idempotency clicked. Pixel-level output validation. Multi-agent architectures with budget-constrained model routing.
2026
Both apps live. Open to what's next. 2,600+ commits across both platforms. The contribution graph tells the story better than any summary. Remote, Chicago area, or hybrid.
02 / What I Ship

Production, not prototypes.

In Production
ColoringStudio.ai
AI Coloring Page Generator

5-path Gemini generation pipeline routing to distinct models. Pixel-level output validation checks monochrome ratio, colorfulness, border continuity, and safe margins before billing. MCP server exposing 5 AI tools via JSON-RPC 2.0. Versioned prompt architecture (v7.1.0) across 21 style families.

React TypeScript Supabase Gemini API Stripe 34 Edge Functions MCP
In Production
StackProof.app
Career Intelligence Platform

3-pass scanner: manifest-only triage via Gemini Flash ($0.01/scan), dual-agent deep scan with chunked batching at 30K-token threshold, tribunal reconciliation merging findings across agents. Budget-constrained model routing with $2.00 per-scan ceiling and 5-minute timeout enforcement.

Next.js 14 PostgreSQL Prisma Claude API Gemini API Multi-Agent
03 / How I Work

Principles over process.

I treat context windows like a kitchen. You don't cook in last night's dirty dishes. Clean session, atomic scope, structured handoff.

Ship first. Refactor when it earns the right. I rewrote billing three times because production taught me what planning couldn't.

Every AI output gets validated before it touches a user. Line-art guardrails, monochrome ratio checks, budget ceilings. Trust the model, verify the output.

The contribution graph doesn't care about your background. It only counts the days you opened your editor.

04 / Get in Touch

Let's talk.

Open to AI Engineering and Full-Stack roles.
Remote, Chicago area, or hybrid.

hello@christianmartin.dev
GitHub ColoringStudio.ai StackProof.app LinkedIn