I spent 5 years running Deal Desks at Adobe, Box, and Otter.ai, managing thousands of enterprise deals, owning Salesforce CPQ end to end, and watching the same workflow bottlenecks slow sales teams down repeatedly.
So I started building the solutions myself.
Today I run Campus Dyno and build two products solo by directing AI coding agents: a multi-tenant CPQ platform and GRAPH/OS, a business-graph platform whose agents produce cited, draft-first actions. Alongside them I run GTM engineering engagements for venture-backed companies, most recently a durable enrichment platform for a healthcare marketplace. Everything I build comes from the same instinct: understand the problem first, then build exactly what the business needs.
- Campus Dyno CPQ: multi-tenant CPQ with document ingestion, bi-directional Salesforce and HubSpot sync, and an NLU quote assistant with deterministic pricing
- GRAPH/OS: versioned business graph with bi-temporal facts and scoped agents producing cited, draft-first actions
- GTM Enrichment Platform: client engagement, a deployed web app holding a 28,811-account prospect universe with dedup as a pre-spend gate
- J-Curve Racing: the commercial-ops platform behind my motorsports business, replacing Shopify, QuickBooks, and a CRM stack
Everything here is production software with a real business on the other end of it, built by one person who understood the problem before writing a single prompt.
If you're looking for someone who bridges domain expertise and product execution, let's talk.
Two products built solo by directing AI coding agents, plus GTM engineering engagements for venture-backed AI companies.
- Multi-tenant CPQ on a TypeScript monorepo: 1,150+ commits and 100K+ lines since March 2026
- Full quote lifecycle behind tenant-isolated auth: deterministic pricing engine, discount-threshold approval routing, framework agreements, branded PDF and Google Doc generation
- Document ingestion: an uploaded order form (PDF, DOCX, or image) becomes a templated quote via PDF parsing, vision OCR, and field-to-schema mapping
- Native bi-directional CRM sync for Salesforce and HubSpot: watermark-based incremental pulls, idempotent upserts, conflict resolution, audit logging
- NLU quote assistant that reads plain-language input and extracts intent; pricing stays in deterministic code, never the model
- AEO content engine and the campusdyno.io marketing site
- Multi-tenant business-graph platform for built-to-order and e-commerce SMBs
- Versioned entity graph: entities, edges, and bi-temporal facts
- Scoped LLM agents produce cited, draft-first actions; a human approves before anything sends
- Postgres row-level security per tenant from day one; pgvector entity resolution across sources
- Source adapters normalize email, forms, commerce, and payment data into one queryable graph
A healthcare marketplace, engaged through Campus Dyno.
- Durable GTM enrichment platform: 14 one-off prospecting scripts consolidated into one tested package covering screening, ICP scoring, persona tiering, deliverability, and cross-list dedup, with 229 tests
- Deployed as a Postgres-backed web app and now the system of record for the client's outbound universe: 28,811 accounts and 1,099 contacts, with upload, dedup, domain resolution, review, and CRM-ready export as one workflow
- Universe dedup as a pre-spend gate: a recent batch added 0 net-new accounts and 5,063 net-new contacts, spending only on the net-new layer
- Pluggable enrichment providers behind one interface: Apollo, ZeroBounce, FullEnrich
- Spend-gated provider dispatch: authoring-time gating, run-time preflight, forecast ceilings, human-approved paid runs
- GTM AI platform ingesting deal data from Salesforce, Slack, DocuSign, email, and meeting transcripts to power a deal-pipeline dashboard and a Slack deal agent
- Built an internal CPQ adopted by 25 users in 50 days; cut quote-creation time about 95% through process redesign and AI-assisted build
- Shipped 8 internal tools across forecast hygiene, partner-deal routing, renewal alerting, ICP scoring, and contract-cleanup automation
- Built a Claude-powered sales-intelligence platform: research-integrity pipeline grounding prospect summaries in primary sources, persona-matched outreach, real-time engagement scoring
- Designed and built end-to-end PLG trial infrastructure: prospect-facing portal, automated token provisioning, AE-facing dashboard, conversion analytics
- Won an internal engineering hackathon, earning a product development green light
- Joined as an SDR; moved into GTM & AI Products after three months. FTE through March 2026, continuing engagement since
- Custom e-commerce and commercial-ops platform: catalog, Stripe checkout and webhooks, a two-axis order state machine, guest checkout, discount codes, admin tooling. Replaced Shopify, QuickBooks, and a CRM stack
- Domain tooling: wheel-pricing calculator with server-side verification, Three.js 3D wheel viewer fed by a photogrammetry pipeline, 71-vehicle fitment knowledge base
- AEO/SEO content engine: 87 published articles and a 100+ prompt answer-engine ranking panel across ChatGPT, Perplexity, Gemini, and Claude
- Founder-led B2B sales end to end: race teams, OEMs, and retail partners
Ran Frame.io's Deal Desk post-acquisition, closing deals to Adobe's quality standards without additional tooling, engineering resources, or headcount. Managed and closed 400+ deals per quarter covering SMB, mid-market, and enterprise across net new, expansion, and renewals.
Founded and built Otter.ai's Deal Desk function from scratch as the sole hire: RFPs, security questionnaires, approval matrix design, legal language library, order management, and a full Stripe-to-Zuora billing migration.
Managed enterprise, mid-market, and SMB deals across the West region and Australia: deal structure, C-level approvals, and net-new, expansion, renewal, and ramp motions.