Most executives talk about AI. For the past two years I've been shipping it — six AI-native products across revenue, security, discoverability, and story, built as a player-coach with AI agents as the team.
A renewals platform for B2B SaaS that treats every renewal as a revenue event — with AI doing the discovery, risk analysis, and prep work an account team would otherwise grind through by hand.
Built around the $500M net-revenue-retention opportunity hiding inside software renewals, and shaped by design-partner engagements with CS and renewals leaders from pre-seed to post-PE.
Headless PAM: containerized hubs on customer networks dial outbound to a broker, so vendors, employees — and AI agents — get least-privilege access to critical systems with full session audit. No inbound firewall rules. No dashboards, by design.
The first PAM built for AI operators. Every capability ships as MCP, CLI, and API surfaces, so an AI agent can request, receive, and use privileged access under policy — with the same audit trail as a human.
Built on nine years scaling SecureLink from $0.1M to $50M ARR (acquired by Imprivata).
Does AI recommend your brand? ARI runs a structured prompt matrix across multiple frontier models and scores brand recommendation 0–100 — the Rotten Tomatoes of AI discoverability.
A multi-model scoring engine (Claude, GPT, Gemini, Perplexity) with snapshot diffing over time and auto-generated audit reports customers actually read.
A marketplace that packages an expert's operating knowledge — playbooks, frameworks, even their calendar — as tools any AI assistant can call. Hire the expert, or hire their brain.
Expert profiles served as dynamic MCP resources over a Neo4j knowledge graph with semantic search, wired to real hiring, messaging, and booking workflows.
A contextual-intelligence layer: shared entity graph, context files, a voice engine, and a fleet of MCP servers (search, think, send, workflows) that every product above plugs into.
One context layer, many products. An entity-first data model means the AI starts from real professional context instead of a blank chat box.
A story-universe builder for young creators — characters, friend groups, rivalries — with an AI companion that keeps the canon consistent and grades each universe for narrative richness.
The AI is a co-author with rules: canon enforcement, relationship inference across friend groups, and a mascot that hands out report cards.
Designed and user-tested with my 13-year-old son.
Every product writes to a shared entity spine — one Supabase instance, schema per product, no data silos.
Capabilities ship as tools an AI can call — MCP, then CLI, then API. Dashboards are the last resort, not the first.
Solo across product, engineering, and GTM — using AI agents as the team. The portfolio is the proof.
Everything on this page runs. Grab twenty minutes and I'll walk you through whichever one fits your problem.