About
I'm an AI developer with a full-stack engineering background. I build agent systems that edit video, a RAG coach that writes scripts, and the SaaS that publishes the results to five platforms.
The full-stack side comes from 8+ years on production enterprise software at J.P. Morgan Chase. I worked on chase.com in React and TypeScript, on Python backends, and on multi-region recovery in AWS. I also led a greenfield React + Django analytics platform from zero. So when I build an AI system, I build the whole product around it: front end, API, data and cloud.
I've spent 15 years writing and directing comedy, and I trained at UCB. I bring that storytelling to how I explain technical work.
Most of my work now is getting models to do real jobs cheaply, and checking that they did them right.
Experience
-
2026 – Now
Lead Engineer · getDropin.io↗
I designed and built DropIn, a multi-tenant SaaS that takes one video from footage to client approval to a live post on five platforms. Front end, API, AI, data and AWS.
- Next.js 16
- Fastify
- Postgres
- Claude
- AWS CDK
-
2017 – 2026
UI Developer & Dev Lead · J.P. Morgan Chase
Built chase.com features at scale. Led a greenfield React + Django analytics platform from zero. Led a multi-region AWS recovery plan, first on the team to adopt AWS ARC. Brought agentic AI workflows into internal dev and ops.
- React
- TypeScript
- Python
- Django
- AWS
- Terraform
Solutions Portfolio
-
Agents · Video
ButterCuts↗
An AI video editor for brands that record on camera. One raw recording becomes a long-form video and Shorts, checked by code and independent AI reviewers before it ships.
- Python
- ffmpeg
- HyperFrames
- Claude
- Codex
-
RAG · Agents
Personal RAG Coach↗
A strategist for brands building a presence online. Turns the teaching of experts a brand trusts into a coach that knows the business and writes ready-to-record scripts.
- FastAPI
- Qdrant
- Voyage
- Opus
- Haiku
-
SaaS · AI
DropIn↗
A video workspace for agencies and brand teams. Client approval to live post on five platforms from one calendar, with copy drafted per platform. Built solo.
- Next.js 16
- Fastify
- Postgres
- BullMQ
- AWS CDK
AI Resources
Prompts and guardrails I use every day. Free to copy.
-
AGENTS.md
Token-efficient agents↗
A drop-in rule file for coding agents. Which model does which job, how to stop paying for waiting, and how to check the cheap models are actually good enough.