Airtable Base Doctor

An Airtable review app that checks schema and sample records, then generates interface previews, redesign proposals, and migration plans.

Built the application across product design, OAuth integration, analysis, data storage, generated previews, and report exports.

The app combines diagnostic checks, interface previews, and redesign reports in one review workflow. It proposes changes without writing them back to Airtable.

Dr. Anne Hussain Platform

A book, blog, and podcast platform with an assistant that retrieves published passages and presents source citations.

Built the content platform, podcast ingestion pipeline, and retrieval assistant as part of a long-term client relationship.

Delivered a content platform with podcast ingestion, source search, and a retrieval assistant. Citations provide material for readers to check; they do not guarantee that a generated answer is correct.

PlatoAI

A philosophy dialogue app with five conversation modes, retrieval from primary texts, and a source panel for checking passages.

Built the application solo, including text embeddings, retrieval, dialogue prompts, speaker coordination, and the Next.js interface.

The application combines text ingestion, retrieval, dialogue prompting, and source inspection. The source panel shows the retrieved context without claiming that every answer faithfully represents it.

TweetNet

A social simulation where configurable AI personas write posts and respond to one another in a shared feed.

Built and operated the application solo, including product design, architecture, frontend, backend, and AI behavior.

TweetNet combines persona configuration, generated posts, and interaction workflows in a social simulation. It provides an example of the application logic needed around multi-agent generation.

CamelCollect.io

A web app and browser extensions for saving LinkedIn posts into organized collections, with onboarding and subscription billing.

Sole engineer across product architecture, app and extension development, billing, admin tooling, deployment, and support.

The product connects browser capture, a signed-in web app, and subscription billing. Error reports and logs help trace problems between them.

AI Adventure

A collaborative storytelling experiment with generated narrative turns, character setup, and branching choices.

Co-built the application in a four-person General Assembly final-project team. My contributions included story-page fixes, usage-limit UI, and character-display improvements.

Delivered an interactive storytelling experiment with generated turns and guided choices.

CoffeeBot

A Discord bot for coffee chats and group coordination, built with a team and awarded Best Discord Bot at GryphHacks 2022.

Co-built CoffeeBot with two teammates. My contribution included core bot functionality, particularly the group-create command.

The team delivered a working Discord bot and received the event's Best Discord Bot award.

DylanOS

A playful browser phone simulation drawn on one p5.js canvas, with generated notifications, gestures, quick settings, and a passcode flow.

Built the canvas renderer, input pipeline, notification generators, and state modules with assistance from an AI coding tool.

A self-contained phone simulation with an accompanying explanation of the rendering, input, state, and performance decisions.

Web apps and product development

Build a customer-facing product or internal tool, including the interface, accounts, data, payments, and integrations it needs.

AI assistants and workflow automation

Help people find answers in your content or complete tasks across your systems. Define what the assistant can access, which actions need approval, and how to evaluate its output.

Codebase review and improvement

Identify bugs, architectural problems, and gaps in testing that make an existing product difficult to change. Prioritize fixes and implement the agreed work.

Full-stack & AI Engineer · DK Solutions

2023 - Present

Build web products, retrieval assistants, and tools that connect to existing workflows.

• Built this site's streaming assistant with server tools, browser actions, and structured project intake
• Built Airtable Base Doctor, connecting read-only OAuth access to schema diagnostics, interface previews, and redesign reports
• Built TweetNet, a social simulation with configurable AI personas and generated posts and interactions
• Work across application code, data models, integrations, deployment, and support

Lead Developer (AI Tooling) · The Shift Weight Loss Program (Dr. Ashley White, MD)

Feb 2025 - Jun 2025

Designed and deployed internal tools for practitioner documentation and program workflows.

• Built internal AI-assisted tools for practitioner workflows
• Worked with healthcare team members on requirements, usability, and handling sensitive information
• Delivered web tools for documentation and program management

Lead Developer · CamelCollect

2023 - Present

Developed the web application and browser extensions for a LinkedIn content collection product.

• Built capture and collection workflows across browser extensions and the web app
• Implemented subscription billing and entitlement checks
• Handled deployment, diagnostics, support, and product changes

Instructional Associate · General Assembly

2022 - 2023

Supported students learning full-stack development through code review, debugging, and project guidance.

• Reviewed student code and helped debug unfamiliar codebases
• Guided students through full-stack architecture and implementation choices
• Explained development fundamentals and maintainable coding practices

Web Developer · Dr. Anne Hussain

Jun 2015 - Present

Built and maintained websites, content platforms, and publishing workflows for Dr. Anne Hussain.

• Built web applications and online learning experiences
• Developed the book, blog, and podcast platform with a retrieval assistant
• Maintained hosting and publishing workflows, including email integrations

IT Head · Samko and Miko Toy Warehouse

Apr 2007 - Apr 2020

Managed IT infrastructure and developed web and reporting tools for retail operations.

• Supported the transition from in-person retail to online sales workflows
• Managed cloud and local infrastructure, security, backups, and recovery practices
• Built reporting and automation for inventory and orders
• Maintained the website backend

Useful over flashy

Start with what people need to do. Use AI where it helps, and simpler code where the rules are already known.

Honest technical judgment

Explain the alternatives, the costs, and the unknowns. Revisit a decision when the evidence changes.

Maintainable after launch

The people maintaining a system need to understand its code, monitor its behavior, and diagnose failures. I plan for that during the build.

Working in an existing codebase

I follow the codebase's conventions, submit changes in reviewable pieces, and explain my reasoning. Teaching and reviewing code at General Assembly helped build that habit.

What is your typical project timeline?

Timing depends on scope, existing systems, integrations, and review needs. After discussing those constraints, we can agree on milestones and a delivery estimate.

Do you offer ongoing support after project completion?

Yes. We can discuss a retainer or separately scoped work for bug fixes, feature additions, and maintenance. Capacity, response expectations, and pricing are agreed before work starts.

What technologies do you work with?

For AI: Anthropic/Claude, OpenAI, the Vercel AI SDK, RAG pipelines, MCP servers, and agent orchestration with tools like Claude Code, Hermes, and OpenClaw. For web: React, Next.js, TypeScript, Node.js, and Python. For infrastructure: AWS, Vercel, and Docker. Full list on the About page.

What does an AI agent project actually look like?

It might help people search your content, assist with an internal task, or coordinate steps across your tools. We define the data it can access, the actions it can take, approval requirements, and test cases before building. Model output still needs evaluation and appropriate review.

Can you set up a personal or team AI agent for us?

Yes. A setup can include Hermes or OpenClaw configuration, custom skills, tool connections, and scheduled tasks. We agree on permissions, approval steps, and how failures are reported. You receive the configuration and documentation at handoff.

Can you review what we're spending on AI?

I can review actual usage and evaluate model selection, caching, context size, and batching. Any proposed changes should be checked against output quality and measured cost. Savings depend on the workload and current setup.

Can I contact you about a full-time role?

Yes. I welcome full-time engineering opportunities as well as freelance projects. Send the job description, a little about the team, and your timeline. We can discuss the role and any hiring logistics directly.

One owner for every write

How one module now owns theme updates from the editor, assistant, shared URLs, and storage, including validation, identity, and targeted DOM writes.

Theme updates: performance baseline

A radius edit went from 54 CSS-variable writes to 1. This note records the measurements, their limits, and the unit tests that now check which variables each operation returns.

Theme identity is determined by history, not by value

A theme edited away from a preset stays custom, even if its values later match. This decision avoids comparing every value against every preset during each colour drag.

Invalid theme values are clamped silently

The editor and AI assistant share one validation policy. The note also corrects its original description of the code's behavior before that policy was implemented.

The anti-FOUC script must be generated, not hand-written

The first-paint theme fix is deferred. When implemented, its script must be generated from the Active Theme module to avoid duplicating validation and CSS-variable mappings.