David Sunnock

Product

Three products in production, one open spec.

I founded Slingshot, a visual editor and deploy layer for AI-generated websites, live at slingshot.build. I built Giftag alone on Monday nights, and it has put more than $5,000 in gift cards into recipients' hands. I built Axiom to replace a $12,000-a-year vendor with a database that costs about $120 a year to run.

Slingshot A visual editor and one-click deploy layer for AI-generated websites. Click any element, change it, publish. The edit lands in your own repo. LIVE IN PRODUCTION

The Slingshot editor open on a coffee shop website, with the primary button selected and its link being edited in a side panel

Axiom A production database that replaced a $12,000-a-year vendor tool and unified member and donor records. IN PRODUCTION $12k → $120 annual cost, before and after Giftag A video greeting card that carries real money. The giver keeps the sentiment, the recipient picks the store. LIVE $5,000+ redeemed by recipients

Operations

Three locations, 2,000+ members, a team of four.

Most of my operations work follows one move: find the expensive dependency, understand the requirement underneath it, and build the leaner system. The four below replaced an unstructured first 60 days for new hires, a paid project tool, a PDF handbook, and the assumption that output only grows with headcount. All of it ran across three locations with a communications team of four.

Employee onboarding A 60-day roadmap for new hires, built with my department. It went well enough that HR is making it the standard for all three locations. BECOMING THE STANDARD 60 days new-hire roadmap, three locations Notion operating system One structured system for how work comes in and gets done: automated deadlines, calendar-aware reviews, story-point capacity. IN USE 1 system replacing email intake and a paid tool Knowledge base A searchable home for every SOP and the employee handbook, replacing a PDF nobody read. IN USE One place for SOPs and the handbook Multi-site communications Content, social, video, and email across three locations, run by two full-time and two part-time people. 2019 → NOW ~2M Instagram views in a year, one coached hire

Results

Budgets down, output up, year over year.

I cut paid spend for the largest campaign of the year by about 90% while attendance set a record. I rebuilt the guest journey and took engagement from under 10% to about 60%. I owned the P&L for a company-owned coffee roaster and grew sales 24%. Recurring subscription spend fell more than 60% on my budget, which went down every year while output rose.

Guest journey Every step from first visit to membership class, mapped and tracked, with drip texts in place of phone calls. IN USE 10% → 60% first-time guests engaged Easter campaign The largest campaign of the year, rebuilt around a hand-picked internal team and sharper creative, at a tenth of the paid spend. RECORD YEAR −90% paid spend, attendance up Battle Creek Coffee Roasters Full P&L ownership of a company-owned roaster: hiring, operations, brand, and growth. P&L OWNER +24% sales, year over year

60%+
cut in recurring subscription spend: 1/10
of quoted cost for in-house 3D renders and studio builds: 6
direct reports across two business units: 15 yrs
one organization, from part-time designer at 16 to the executive team

The one that failed

One company wound down. Fifteen mistakes written up.

Find A Farm had a working product in two countries and still failed. The debrief we wrote in January 2021 is the most useful document the company produced. The lesson in it about non-technical users needing tools built for them is the reason I built Slingshot.

Find A Farm A marketplace for buying direct from farms, across two countries. Wound down and debriefed in full. → JAN 2021 15 mistakes named, five things that worked

The lesson that stuck Farms found bank setup hard, the UX hard to navigate, and onboarding too thin. Every tool I have built since starts with the non-technical user.

How I work

Recommend first, then decide. Verify, don't trust.

I build through a human-in-the-loop AI pipeline. One model acts as a technical product manager and pressure-tests the plan. Another implements it against the real codebase. I own direction, tradeoffs, and verification, and I check the deployed commit before calling anything done. Three rules sit underneath all of it.

01

Ship the smallest real thing

Giftag's first milestone was a recipient scanning a code and redeeming a real gift card, not a feature list. Slingshot shipped its first phase behind a 14-check smoke test of the full editor.

02

Write the debrief

When Find A Farm wound down, we listed fifteen mistakes and five things that went right. The product worked. The plan around it did not, and we wrote down why.

03

Build for the non-developer

Find A Farm lost partly because farms could not get through onboarding. Slingshot's editor is built so a non-developer can change the text without breaking anything.

What I bring

Operator, product builder, and comfortable owning the P&L.

Pulled from the projects above, not from a job description.

Operations and systems

Process and workflow design
Intake architecture and automation
Capacity and bandwidth modeling
Vendor and tech-stack rationalization

Product and engineering

Internal tools, spec to maintenance
Database design on Supabase
Build-versus-buy analysis
Human-in-the-loop AI development

Growth

Budget and spend ROI
Paid and organic channel mix
Funnel and lifecycle design
CRM and messaging automation

Leadership

P&L ownership
Hiring, coaching, team management
Multi-site management
Getting frameworks adopted by other departments

I find gaps and opportunities, then build the fix.