AI & Digital / built systems

AI & Digital.

Systems I have designed and built around real work: CreatorOps for creator-support operations, and this portfolio for structured content, publishing and deployment.

01 CreatorOps02 Portfolio platform
Build what the work is missing

Practical problems turned into working systems I can use, test and keep improving.

01 / Operational AI

CreatorOps.

A working agentic operations layer for high-volume creator support. CreatorOps turns long Discord conversations into persistent, searchable cases with evidence, status and clear next actions, then prepares the work needed to respond, escalate or update creators.

2,800+Completed case records indexed
1,200+Records of support handled by me
Operating principle

Read-only ingestion. Evidence-backed history. Agent-prepared actions and responses.

Operational flow
01Discord
02Cases
03Evidence
04Action-ready
Problem

Important creator issues can disappear inside fast-moving support conversations and repeated questions.

System

CreatorOps reconciles new reports against history, links the underlying evidence and keeps a durable case record.

Automation path

The current deployment holds outbound creator messages for review. The same action layer is designed to extend into automatic responses, creator updates and routine follow-up where that level of automation is appropriate.

RailwaySupabaseVercel / Next.jsGitHubKnowledge retrievalEvidence-linked records
View full CreatorOps case study ↗
02 / Web platform

This portfolio.

More than a set of pages. The site is a structured publishing system for professional experience, creator work, performance data and a large photography archive, with deployment and recovery workflows built around it.

What the system handles

Content is separated from presentation so projects, campaign evidence, professional case studies and photography can stay maintainable instead of being hard-coded as one-off pages.

01Content
02Structured data
03Build
04Deploy
616Images in the public photography archive
LiveProduction deployment with release checks
Next.jsReactVercelSupabaseCloudinaryStructured content
View portfolio build case study ↗
AI & Digital

The common thread is simple: identify the repeated friction, structure it, then build something useful enough to become part of the workflow.

Discuss a project ↗