AI Engineer
Built the Eboost Creative Operations and Intelligence Platform. Owned the UI, PDF reports, approval workflow, and AI investigation architecture; collaborated on backend and advertising data pipelines.
- Reduced monthly report preparation from three days to 30 minutes per brand with Creative Audit, which retrieves BigQuery results and drafts AI analysis for human review. Onboarded eight brands.
- Built client ad approvals with decision records and email notifications; imported and verified 1,225 historical ads for a client handling 25 to 35 new ads per month.
- Cached initial creative analyses for reuse across audits, reducing repeated video analysis and token usage. Campaign managers used the recommendations to plan future creatives.
- Built Software Factory, a Claude Code plugin connecting Asana tickets to isolated Git worktrees, headless builders, independent review, CI gates, deployment verification, and ticket closure.
- Implemented AI investigation architecture using LangChain for model interactions and retrieval, and LangGraph to coordinate a supervisor, performance analyst, creative analyst, and evidence reviewer.
- Used Claude/OpenAI APIs, function calling, MCP tools, and brand-context RAG. Built evaluations of numbers, citations, brand fit, output quality, failure rates, latency, and cost, with human approval of reports.
- Built the interface with React, Next.js, and TypeScript. Checked staging deployments on Vercel and Cloud Run with Neon; deployed approvals in Docker on a Hostinger VPS.
- Tracked usage with PostHog and monitored logs, health checks, uptime, and failure alerts. Tuned model cost, response time, validation, caching, timeouts, and notification timing.