I help founders turn strategy into execution. In 9 years as a CEO's strategic right hand, I've built the systems, cadence, and visibility that drive consistent outcomes across sales, partnerships, and operations.
AI is my leverage layer, not a novelty.
I start with problem framing and clear specs — not vague goals. Then I build the system: the operating cadence, the dashboards, the workflows, the follow-up loops. I create simple, durable views so leadership can see what's working, what's stuck, and what needs attention. And I use AI and automation as leverage to make the whole thing self-sustaining — built once, reused often, designed so outcomes don't rely on individual heroics.
Every system below was taken from concept to live output — most in days, not weeks. AI and automation are the execution layer, used to compress time-to-impact and increase quality across every build.
System: Custom Claude skill that pulls both HubSpot pipelines (Partner Efforts + Sales), cross-references Monday.com commitments, categorizes deals by urgency, flags stale activity, and offers to draft follow-ups — all from a single trigger phrase.
Leverage: Claude reads live CRM data via MCP connectors, cross-references a second platform, and generates a prioritized action list with draft outreach ready to send.
Judgment: Designed the triage logic — what counts as "stale" vs. "dead," which deals need me vs. leadership, and how to surface the 3 things that actually matter out of 40+ open items. The AI fetches; I decide.
System: Six Zapier-powered automations covering the full deal lifecycle: Stale Deal Alert, Close Date Approaching, Overdue Close Date, Missing Next Step, CEO Task Nudger (for a CEO who doesn't use the CRM), and New Lead Speed Alert. All post to Slack with direct HubSpot links.
Leverage: Claude designed every trigger condition, message template, and edge case handler from plain-English problem descriptions. No code written — but every automation reflects real operational logic, not generic templates.
Judgment: The CEO Task Nudger is the standout: it solves a behavior problem, not a tool problem. Instead of building another dashboard the CEO won't check, I automated accountability into a Slack channel he already lives in. AI didn't choose that — I did.
System: Custom Claude skill that searches conversation history for OKRs, active priorities, and open commitments, then formats a structured daily briefing: Top 3 Priorities, Meeting Prep, Predictable Profits Tasks, Personal/Development, and Urgent/Overdue flags.
Leverage: Claude searches across weeks of conversation context, synthesizes current state across multiple workstreams, and formats output without requiring re-explanation of any project.
Judgment: Built because the first 30 minutes of every day were spent reconstructing context. The skill itself isn't technically impressive — what matters is identifying that the problem existed, scoping to exact need, and building with zero bloat. Best tools are invisible.
System: Upload a resume, get tailored STAR answers for 10 behavioral questions. Inline editing, self-assessment rubric, model examples, one-click export.
Leverage: Claude generates all STAR content, powers refinement, and built the entire app from spec to deployment in 3 days.
Judgment: Designed the UX based on my own interview prep pain points. Scoped features ruthlessly — no auth, no database, just the core loop.
System: Simulates a Series A board with P&L responsibility. Three NPCs with distinct priorities challenge your decisions. Monthly financials, delayed consequences, commitment tracking.
Leverage: 100% Claude-powered — simultaneously roleplays multiple characters, generates consistent SaaS financials, and maintains state across time skips.
Judgment: Designed the scenario architecture so decisions have delayed, compounding consequences — not instant feedback. That's what makes it feel real.
System: Extracted structured data from 80 unanalyzed Word docs, cross-referenced against health dashboard and attrition records, and generated a leadership report with churn predictors.
Leverage: Claude designed extraction patterns and ran statistical analysis — work that would've taken days manually.
Judgment: Caught a base-rate fallacy in the AI's output. The initial analysis was wrong; I forced a corrected conclusion before presenting to leadership.
System: Six-question quiz → customized pipeline with stage definitions → professional PDF → one-click deploy to HubSpot.
Leverage: Claude built the full app, designed pipeline logic from sales methodology, implemented OAuth, and debugged API constraints.
Judgment: Discovered free-tier API constraints mid-build and designed a workaround. Scoped a lead capture mechanism into the flow — not in the original brief.
System: Fetches audio from Slack → Whisper transcription → Claude thematic analysis → posts structured summary back to Slack channel automatically.
Leverage: Multi-AI pipeline — Whisper transcribes, Claude analyzes and structures. Runs at $0.36/mo.
Judgment: Identified a dying practice and bet that consumable output — not better compliance — was the fix. The team started recording again within a week.
System: End-to-end NPS pipeline — pulls clients from HubSpot, sends surveys via Gmail, routes alerts to Slack by score tier, logs everything to a spreadsheet.
Leverage: Claude wrote all 400+ lines of Apps Script and designed the routing logic. I described the system; it built it.
Judgment: Chose the right constraint (zero budget) and scoped the MVP. Decided what to automate vs. what to leave manual.
System: Tier-based churn analysis across four programs — tenure comparison, challenge evolution mapping, and retention pattern identification. Delivered as a leadership briefing.
Leverage: Claude performed all statistical analysis and pattern identification across the dataset.
Judgment: Challenged an incorrect segment conclusion from the AI. The initial output would have misled leadership — I caught it and forced a corrected analysis.
System: Fetches partner channel messages → Claude parses unstructured text for registrations, attendees, conversions, and revenue → batch-writes to Google Sheet.
Leverage: Claude API serves as the parsing engine — extracting 4 structured metrics from ~37 unstructured Slack messages per run.
Judgment: Recognized that the bottleneck wasn't the spreadsheet — it was the 15 minutes of reading and interpreting Slack messages. Automated the hard part, not just the data entry.
Every system has a story. These are the ones that didn't make the carousel but still shipped.
The most important thing I do isn't building systems — it's deciding which system to build next. At any given time there are fifty things that could be improved. My job is figuring out which three will actually move the business, and ignoring the rest. That judgment comes from 9 years of being close enough to the CEO to understand what matters, and close enough to the work to know what's realistic.
An example: when I analyzed 80 client intake documents, the AI flagged a pattern that looked like a breakthrough insight. But the math was wrong — a base-rate fallacy that would have sent leadership chasing a ghost. I caught it, corrected the analysis, and delivered findings that actually held up. That's the difference between using AI and being useful with AI.
Before this, I co-founded a healthcare technology startup focused on prescription safety that was acquired in 2016. I'm currently finishing my B.A. in Communication Studies at Sacramento State. Outside work, I climb rocks and play basketball — physical challenge is how I stay sharp.
I'm seeking a Chief of Staff or Strategy & Ops role where I can be the person who turns a CEO's hardest problems into working systems. If that's what your team needs, let's talk.