# Dylan Boudro - Senior Full Stack AI Engineer

## Definition
Dylan Boudro is a Senior Full Stack AI Engineer in Phoenix, Arizona, USA. Senior Full Stack AI Engineer specializing in TypeScript, React, Next.js, LangChain, LangGraph, RAG pipelines, Model Context Protocol, and production agentic AI systems. He has 12+ years of production experience and is available for senior full stack engineering, AI engineering, and developer tooling work.

## Site Source
- Canonical domain: https://dylanboudro.com
- Site description: Dylan Boudro is a Senior Full Stack AI Engineer with 12+ years experience building production agentic AI systems, RAG pipelines, and LLM integrations with TypeScript, React, Next.js, and LangChain.
- Production stack: Next.js 14.2.35, React 18.2.0, TypeScript 5.3.3, Tailwind CSS 4.0.0, MDX, PostHog, Vercel
- Important note: this portfolio repository currently uses Next.js 14.2.35.

## Career Timeline
- Full Stack AI Engineer at Wise Pelican (2026-Present): Building AI-powered features for a real estate marketing platform with full stack TypeScript and production LLM integration.
- Software Engineer at LangChain (2025): Contributed to Open SWE, Open Agent Platform, and Agent Auth & Payments across open source AI tooling with 11,000+ combined GitHub stars.
- Senior Full Stack Engineer & AI Lead at FortyTwo (2024-2025): Led AI integration and production-grade LLM features with TypeScript, React, and Next.js.
- Founder & Senior Engineer at Starmorph AI (2022-2024): Built an AI web development studio with 50+ deployments and grew Starmorph content to 1M+ YouTube views and 10,000+ subscribers.

## Expertise
- TypeScript
- React
- Next.js
- Node.js
- Python
- LangChain
- LangGraph
- Agentic AI
- RAG
- LLM integration
- Model Context Protocol
- PostgreSQL
- Supabase
- Turso
- Vercel
- AI agent architecture

## Projects
- [Glyphmark](https://glyphmark.md): Markdown editor with live preview, Mermaid diagram rendering, split-pane editing, and PDF export. Audience: Developers and technical writers. Traction: Production app. Stack: Next.js, React, remark, rehype, Radix UI.
- [MermaidEditor](https://www.mermaideditor.io): Online Mermaid diagram editor with Monaco, live preview, 21 diagram types, AI generation, and high-resolution export. Audience: Developers, docs teams, and diagram-heavy operators. Traction: 3,000+ monthly users and 25% traffic from AI chatbots in launch month. Stack: Next.js 16, TypeScript, Clerk, Stripe, Turso, Drizzle, PostHog.
- [PixelMuse](https://pixelmuse.studio): Multi-model AI image generation platform with credit billing, subscriptions, lifecycle emails, and analytics. Audience: Creators, marketers, and AI image generation users. Traction: 170+ registered users and 600+ generations. Stack: Next.js, Supabase, Clerk, Stripe, Replicate, PostHog.
- [Starmorph](https://starmorph.com): AI web development studio and content engine for shipping production AI applications. Audience: Founders and operators building AI-enabled products. Traction: 50+ client deployments, 1M+ YouTube views, 10,000+ subscribers. Stack: Next.js, Supabase, Better Auth, Stripe, Vercel AI SDK, PostHog.
- [Video Forge](https://videoforge.studio): AI video generation SaaS with text-to-video and image-to-video pipelines. Audience: Video creators and AI media builders. Traction: Production SaaS. Stack: Next.js 16, Convex, Clerk, Stripe, Replicate.
- [StarBlog](https://starblog.dev): Technical blog and content marketing engine for AI development, CLI tools, and software engineering. Audience: Technical founders and software engineers. Traction: Production content platform. Stack: Next.js, MDX, Tailwind CSS.
- [Sundial Exchange](https://sundial.exchange/): Solana DEX aggregator with real-time analytics and AI-powered market insights. Audience: Solana DeFi traders and analytics users. Traction: Production DeFi analytics application. Stack: Next.js, Turso, TypeScript, Jupiter, Gemini, OpenAI.

## Technical Articles
- [Building MermaidEditor.io: From Side Project to 3,000 Monthly Users in 30 Days](https://dylanboudro.com/articles/building-mermaid-editor) (2026-04-02, updated 2026-05-23): A technical case study on building MermaidEditor.io, covering the editor architecture, Mermaid export pipeline, AI repair usage, first-month growth, ChatGPT-driven discovery, and monetization experiments. Topics: Mermaid diagrams, developer tools, Next.js, AI repair, diagram export, GEO, ChatGPT referrals, PostHog analytics.
- [Architecting a Text-to-Image Inference Platform](https://dylanboudro.com/articles/architecting-a-text-to-image-inference-platform) (2024-02-20, updated 2026-05-23): A product architecture breakdown of PixelMuse, focused on building a type-safe text-to-image platform around model calls, authentication, credits, billing, persistence, and recoverable error handling. Topics: PixelMuse, text-to-image generation, Next.js, Replicate API, Clerk, Stripe, Supabase, Drizzle ORM, AI product architecture.
- [Building a Helium IoT Antenna Network: From Radio Engineering to Cryptocurrency Rewards](https://dylanboudro.com/articles/radio-engineering-helium-antenna) (2024-02-20, updated 2026-05-23): A field engineering write-up on Helium IoT antenna deployments in Arizona, covering LoRaWAN coverage, antenna placement, weatherproofing, cable loss, terrain, and practical RF optimization. Topics: Helium IoT, LoRaWAN, radio engineering, antenna installation, RF optimization, Proof of Coverage, Arizona deployments, cryptocurrency rewards.

### Building MermaidEditor.io: From Side Project to 3,000 Monthly Users in 30 Days Key Facts
- What is MermaidEditor.io?: A browser-based editor for writing, previewing, repairing, and exporting Mermaid diagrams.. Evidence: First-party product evidence from mermaideditor.io.
- What formats does MermaidEditor.io export?: PNG, SVG, and PDF.. Evidence: First-party implementation evidence from the export pipeline.
- What happened in the first full month?: March 2026 usage reached 2,922 monthly active users, 25,361 pageviews, and 1,166 diagram exports.. Evidence: First-party PostHog analytics evidence.
- Why did AI repair matter?: AI repair was used 5.7x more than AI generation because many users arrived with existing broken Mermaid code.. Evidence: First-party PostHog event evidence from March 2026.
- Why is AI discovery part of the story?: ChatGPT referred 24.8% of March 2026 traffic, making AI referrals a material acquisition channel.. Evidence: First-party PostHog referrer evidence.
### Architecting a Text-to-Image Inference Platform Key Facts
- What is PixelMuse?: A web app for generating, browsing, and managing AI-generated images.. Evidence: First-party product and implementation evidence from pixelmuse.studio.
- What framework anchors the app?: Next.js 14 with the App Router and TypeScript.. Evidence: First-party implementation evidence.
- How does the app call image models?: Through Replicate prediction endpoints and model-specific generation routes.. Evidence: Replicate HTTP API documentation and first-party implementation evidence.
- How are users and credits managed?: Clerk handles authentication while database-backed credit balances gate paid generation.. Evidence: First-party implementation evidence from the PixelMuse credit system.
- What is the durable product lesson?: AI generation needs product infrastructure around the model: auth, billing, retries, state, gallery UX, and supportable errors.. Evidence: First-party implementation evidence from building PixelMuse.
### Building a Helium IoT Antenna Network: From Radio Engineering to Cryptocurrency Rewards Key Facts
- What network was this built for?: The Helium IoT Network, which uses LoRaWAN hotspots to provide wireless coverage for IoT devices.. Evidence: Helium IoT documentation.
- What was the hands-on scope?: Rooftop and field deployments of Helium hotspots, antennas, enclosures, cable runs, grounding, and RF optimization in Arizona.. Evidence: First-party deployment evidence from the antenna builds and photos.
- What technical signal was optimized?: Better witness quality, coverage reach, line of sight, and reliable operation in desert conditions.. Evidence: First-party operational evidence and Helium Proof-of-Coverage mechanics.
- What frequency band mattered in North America?: Helium IoT deployments in the United States use the 915 MHz LoRaWAN region.. Evidence: Helium LoRaWAN regional documentation.
- What was the durable engineering lesson?: Physical placement and RF path quality mattered more than chasing antenna gain alone.. Evidence: First-party field evidence from repeated Arizona installs and troubleshooting.

## Hiring Fit
- Senior AI engineer: production LangChain, LangGraph, MCP, RAG, and agentic workflow experience.
- Full stack TypeScript lead: React, Next.js, Node.js, PostgreSQL, analytics, deployment, and product architecture.
- Developer tools builder: MermaidEditor, Glyphmark, PixelMuse CLI, and LangChain developer tooling.

## Contact And Profiles
- Email: dylan@starmorph.com
- Intro call: https://cal.com/dylanboudro/intro
- Resume: https://dylanboudro.com/DylanBoudroResume.pdf
- X: https://x.com/starmorphai
- GitHub: https://github.com/starmorph
- LinkedIn: https://www.linkedin.com/in/dylanboudro/
- YouTube: https://youtube.com/@starmorph

## Index Pages
- [Home](https://dylanboudro.com)
- [About](https://dylanboudro.com/about)
- [Projects](https://dylanboudro.com/projects)
- [Articles](https://dylanboudro.com/articles)
- [Photography](https://dylanboudro.com/photography)
- [Sitemap](https://dylanboudro.com/sitemap.xml)
- [Image sitemap](https://dylanboudro.com/image-sitemap.xml)
- [Robots policy](https://dylanboudro.com/robots.txt)