PROJECTS.SHEET

Lundstrum — AI ops + living institutional site

CS / LUNDSTRUM-AI-OPS

Six-agent operations system on a 200+ file corpus, plus Remotion/generative video for donor fundraising.

Role
AI Systems Consultant — Lundstrum Performing Arts
Period
Jan – May 2026
Stack
Next.js 15 · React 19 · TypeScript · Tailwind · Remotion · Claude Code

Visual artifact

Lundstrum case study PDF
Candidate-kit case study PDF

Problem

A nonprofit arts organization needed operational leverage (marketing, grants, donor comms, events) grounded in its real institutional knowledge — not generic AI demos.

Approach

Architect specialized AI agents with a shared 200+ file corpus, custom skills for on-brand UI generation, and a Remotion + generative-API video pipeline for live fundraising.

Decisions

  • Six-agent autonomous ops spanning marketing, grants, donors, events, analytics, and content (core-resume.md).
  • Corpus ingest of website pages, blog archive, annual report, and rehearsal transcripts for grounded grant and content work.
  • Remotion programmatic video integrated with Higgsfield AI and Luma Labs for a live donor campaign.

Outcome

Documented in core-resume.md: six-agent system, 200+ file corpus, custom skills/workflows, and Remotion fundraising pipeline. NO FABRICATED TRAFFIC/DONATION LIFT — not claimed here.

What I'd do differently

Would publish a short agent-eval scorecard next to the six-agent diagram so interviewers can see failure modes and handoff rules, not only the happy path.

Sources logged for review — see PR description for human flags.