Operations · Selected Work
I take ownership of how it runs, then hand it back stronger.
A full operations lead, part of the time. I map how work actually flows, find what is quietly costing you time and money, fix it, and build the systems so it stays fixed after I stop showing up every week.
Ellyn Shirk
I run my own company as a one-person operation on systems I built myself. That is the same thing I would be doing for you.
Case Studies
Systems I built and still run.
Not slide decks about operations. Working systems, against live data, that keep running when nobody is watching them.
A company run by one person and a stack of systems
Everlight is my products business. I am the only employee, which meant I either accepted a ceiling on what one person can hold in their head, or I built the operating layer that holds it instead. I built it.
It runs on an AI chief of staff that briefs me every morning on what actually needs attention, pulled from live data rather than a static checklist. Around that sits an advisory panel I put real decisions in front of before committing, and content pipelines that turn one input into finished on-brand assets.
The point is not that I use AI. Plenty of people use AI. The point is that this is a functioning operating system for a real company with real money moving through it, and it has been running without me babysitting it.
Add before sending
One screenshot of the morning briefing does more here than three paragraphs. Reseed with invented data first. The rule is not "does it say a client name," it is "is there any real client data on this screen."
What this proves
- I design the whole operating layer, not one clever tool
- I wire systems to live data and keep them running
- I know what a small business actually needs because I run one
- I automate the repetitive half so judgment gets the attention
The tools most teams rent, badly, for more than they cost to build
Most small teams are paying monthly for four tools that each solve sixty percent of their problem, with a person quietly gluing the gaps together by hand. I build the thing that fits instead.
That means a CRM shaped like how you actually sell rather than how a vendor assumed you do, a live dashboard that answers the questions you keep asking in Slack, document systems that produce on-brand work in seconds instead of an afternoon, and pipelines that move data between the places it needs to be without anyone remembering to do it.
I would rather show you than describe it. Ask, and I will walk you through one working end to end instead of handing you a screenshot of it.
Add before sending
Two or three screenshots: the CRM, a dashboard, a generated document. Every one must be seeded with invented companies and people before it goes on this page. Build the demo dataset once and reuse it everywhere.
What this proves
- I build the tool rather than subscribe to one that half fits
- I work across data, interface, and the automation between them
- I reduce the number of things a small team has to pay for
- I can show working software, not a specification
Nine products, one person, and the reason that is possible
I am building a suite of nine products on my own. The first one is live. The interesting part is not any single product. It is that the ninth will cost me a fraction of what the first one did.
That happens on purpose. One roadmap that every decision gets written back into, so nothing has to be re-litigated six weeks later. A shared design system, so every product ships looking like a sibling instead of a stranger. Build patterns reused rather than reinvented, so the boring eighty percent is already answered before I start.
This is the actual work of operations, and it is the part most businesses never get to. Not the heroic push to ship one thing. The unsexy discipline that makes the next thing cheaper than the last one.
What this proves
- I standardize so the second time costs less than the first
- I keep one source of truth instead of decisions scattered across chats
- I ship repeatedly, without a team behind me
- I build for reuse, which is where the compounding is
How It Works
Audit, build, stay.
Low commitment at the start, because you have no reason to trust me yet. The part that matters most is the third one.
A two-week audit
I learn the business properly, map how work really flows, and rank the few things costing you the most. You keep the plan whether or not we continue.
Fixed feeWe work the plan
I rebuild the processes, stand up the systems, and put automation on the repetitive work. You feel the business get lighter month over month.
Month to monthI stay and keep it running
Systems drift. Tools change, integrations break, and the process everyone agreed to erodes quietly. I stay on at a reduced scope so what we built still works next year.
Ongoing, smallerWhat I Do
Where I am actually useful.
Industry does not matter much. Readiness does. If everything still runs through you, this is the work.
Process and documentation
Repeatable workflows for onboarding, delivery, and the daily machine, so nothing depends on you remembering it.
Systems and tooling
A CRM that fits how you work, a live dashboard of business health, and document systems that produce on-brand work in seconds.
AI that earns its keep
Automation on drafting, summarizing, follow-ups, and briefings. Fewer hours spent on work a system should be handling.
Standardization
Templates, patterns, and a single source of truth, so the second time you do something costs a fraction of the first.
Next Step
Tell me what keeps breaking.
You probably do not need a full-time operations hire. You need someone to build the thing properly and then stay to keep it running.
Start with the audit. Two weeks, a plan you keep either way, and an honest answer on whether ongoing help makes sense. If it does not, I will tell you that.