12+ years owning strategy, structure and budget for multi-million-dollar paid engines — $40M annual budget authority, media buying scaled from POC to $20M/year, teams of 5–15. I don't just run the spend; I build the systems underneath it. The four platforms below are live production builds — architected, shipped and operated end-to-end, solo, with an AI-native toolchain. Every demo is fully interactive.
Faithful, fully functional replicas of the live products, running on generated sample data — click anything. The production versions run with real revenue and customer data behind auth. Each demo opens in a new tab.
A programmatic buy/sell operations cockpit: reconciles reporting from every ad platform (display, video, CTV, in-app, push) into a single contribution-margin view and converts it into ranked spend actions with trust-gated automation. ClickHouse analytical spine, restatement-safe fact tables, JSON adapter ingest, scheduled reporting.
A server-to-server tracking & attribution network: click/conversion tracking, per-geo and revshare pricing, publisher postbacks with retry sweeps, and a role-scoped reporting portal with fail-closed permissions. Fraud handling built in at the edge: invalid-traffic screening (1,500-pattern crawler corpus, datacenter-ASN detection), incentivised-traffic flagging, bot/test segregation, and click-domain burn-switch failover.
The internal revenue system of record for my consultancy: partners, six-stage weighted pipeline, invoices & insertion orders, a Business/Corp P&L workbench with approval-driven payment tracking, an ads.txt-derived market prospecting board, live read-only Xero reconciliation and Wise treasury — plus an AI layer that parses forwarded partner emails into proposed deal updates a human approves.
A cross-platform consumer app (Firebase, Capacitor iOS/Android, secure document vault): an AI-first relocation assistant where "Kate" interviews you, generates your whole plan, briefs you daily, and proposes concrete changes you approve with a tap. Built on the principle: an assistant, not a liability.
My career is the revenue side of paid media: Google, Meta, Taboola, Outbrain, TikTok, and RTB/programmatic across display, video, CTV, in-app and push. I've scaled a new revenue stream from POC to $1M+/month, grown a publisher network 400% to secure $2.5M/month, and served as company-wide subject-matter expert in media buying while leading multidisciplinary teams.
The difference: I personally build the machinery under the spend — automated buying infrastructure, pause/scale/budget playbooks, guardrails and alerting, restatement-safe data pipelines. One operating principle runs through everything: allocate capital on true margin after fees and returns, never on platform-reported ROAS alone.
AI-first since founding Vane Labs: Claude Code and agentic workflows in daily production use — which is how one person ships and operates the four platforms above in months, not years.
If you need someone who can own a paid-media engine at scale and build the AI-first machinery behind it, I'd love to chat.