What I Know to Be True as Chief of Staff at an AI-Native Startup
Plan your expenses and deductions strategically each year
Egestas tincidunt ipsum in leo suspendisse turpis ultrices blandit augue eu amet vitae morbi egestas sed sem cras accumsan ipsum suscipit duis molestie elit libero malesuada lorem ut netus sagittis lacus pellentesque viverra velit cursus sapien sed iaculis cras at egestas duis maecenas nibh suscipit duis litum molestie elit libero malesuada lorem curabitur diam eros.
- Morbi fringilla molestie magna sed dictum. Praesent pharetra turpis augue.
- Cras mi purus, viverra vitae felis sit amet, tincidunt fringilla lorem.
- non mattis urna ex nec sem. Donec varius diam et suscipit venenati proin tincidunt
- Quisque euismod posuere lacus sit amet volutpat. Praesent vel imperdiet
Maintain accurate and organized financial records for easy reporting
Tincidunt pharetra at nec morbi senectus ut in lorem senectus nunc felis ipsum vulputate enim gravida ipsum amet lacus habitasse eget tristique nam molestie et in risus sed fermentum neque elit eu diam donec vitae ultricies nec urna cras congue et arcu nunc aliquam at.
Talk to a tax professional for expert guidance
At mattis sit fusce mattis amet sagittis egestas ipsum nunc scelerisque id pulvinar sit viverra euismod. Metus ac elementum libero arcu pellentesque magna lacus duis viverra pharetra phasellus eget orci vitae ullamcorper viverra sed accumsan elit adipiscing dignissim nullam facilisis aenean tincidunt elit. Non rhoncus ut felis vitae massa mi ornare et elit. In dapibus.
- Morbi fringilla molestie magna sed dictum. Praesent pharetra turpis augue.
- Cras mi purus, viverra vitae felis sit amet, tincidunt fringilla lorem.
- non mattis urna ex nec sem. Donec varius diam et suscipit venenati proin tincidunt
- Quisque euismod posuere lacus sit amet volutpat. Praesent vel imperdiet
Personalized Marketing: AI helps in analyzing customer data to create personalized marketing campaigns
At mattis sit fusce mattis amet sagittis egestas ipsum nunc. Scelerisque id pulvinar sit viverra euismod. Metus ac elementum libero arcu pellentesque magna lacus duis viverra. Pharetra phasellus eget orci vitae ullamcorper viverra sed accumsan. Elit adipiscing dignissim nullam facilisis aenean tincidunt elit. Non rhoncus ut felis vitae massa. Elementum elit ipsum tellus hac mi ornare et elit. In dapibus.
“Amet pretium consectetur dui aliquam. Nisi quam facilisi consequat felis sit elit dapibus ipsum nullam est libero pulvinar purus et risus facilisis”
Use accounting and tax software to make management easier
Placerat dui faucibus non accumsan interdum auctor semper consequat vitae egestas malesuada quam aliquam est ultrices enim tristique facilisis est pellentesque lectus ac arcu bibendum urna nisl pharetra bibendum felis senectus dolor commodo quam elementum sapien suscipit qat non elit sagittis aliquam a cursus praesent diam lectus tellus mi lobortis in amet ac imperdiet feugiat tristique nulla eros mauris id aenean a sagittis et pellentesque integer ultricies sit non habitant in cras posuere dolor fames.
Chief of Staff is a strange seat. You’re not the CEO, you’re not an engineer, and you’re not the one closing the books. You sit in the middle of everything, which means you get to watch how an AI-native company actually runs — not how LinkedIn says it runs.
I have that seat at Till CFO, an embedded finance team built by operators on an AI-native stack. Our accountants open a terminal before they open their email. I’m not an engineer either — my background is operations, people, and an unreasonable number of spreadsheets. So consider this field notes from the middle of the org chart.
Here’s what I know to be true.
1. AI Fluency Is a Reps Problem, Not a Tools Problem
Tool demos are exciting for exactly one afternoon. Someone shares their screen, the room says “whoa,” and two weeks later nobody can reproduce what they saw. Watching someone else use AI is like watching someone else go to the gym. Genuinely nothing happens to you.
What works is repetition with structure: small daily challenges, everyone participates, no exceptions. Some days it takes ten minutes. Fluency is a motor skill, and motor skills don’t care how inspiring the kickoff meeting was. They care whether you showed up today.
One rule matters more than the rest: nobody gets left behind. The moment AI skills become optional, they become a side project for the already-curious, and the gap between your most and least fluent people turns into a canyon.
2. The Team Is the Hero. AI Is the Plumbing
Every AI-native company eventually faces the same design question: where do the people go?
The wrong answer is “wherever the AI can’t reach yet.” The right answer is “wherever the judgment is.”
We ran this on our own books. We let AI reconcile half a million dollars of our own cash — real deposits, real invoices, real general ledger. When the work was done, exactly six decisions needed a human: a renewal term, a pricing structure, how to treat a client paying under a name our systems didn’t recognize. The machine did the gathering. People owned the calls.
That ratio is the whole thesis. AI absorbs the volume. The humans get the decisions. If your AI rollout makes your people feel smaller, you’ve built it backwards.
3. Curiosity Is the Qualification
The people who got fluent fastest at Till weren’t the youngest and weren’t the most technical. They were the most curious. The teammate who asks “wait, could it also do this?” laps the teammate with the computer science background who already decided what AI can’t do.
That changed how I think about hiring. Resumes tell you what someone has done. They’re quiet on whether someone will poke at a new capability until it gives. We hire for the poking.
4. If It Isn’t Written Down, It Doesn’t Exist
AI can only use knowledge it can read. The deal that changed pricing mid-quarter, the client that pays under a different legal name, the reason an invoice looks irregular — in most companies, that context lives in someone’s head, and it walks out the door at 5pm.
At an AI-native company, writing things down stops being hygiene and becomes infrastructure. Our systems cite their sources because the sources exist to cite. The discipline is boring. The compounding is not.
5. Every Process Will Be Rebuilt Within a Year. Document It Anyway
This one took me a while to make peace with. You write the runbook, the capability shifts, and the runbook is wrong by spring.
Document it anyway. The rewrite is fast precisely because the old version exists. Teams that skip documentation because “it’ll change anyway” don’t end up agile. They end up dependent on whoever was in the room.
6. You Can’t Sell What You Don’t Live
Till sells embedded finance teams: senior operators wrapped around your business, running on the same AI-native infrastructure we run ourselves. And the reason that pitch lands isn’t the deck. It’s that everything in it happened to us first.
The messy system migration. The cash that fell behind. The reconciliation that AI did in hours instead of weeks, with a human making exactly the calls a human should make. We are our own first client, every time. When a company runs its own operations on the thing it sells, the gap between marketing and reality goes to zero. That’s rarer than it should be.
The Honest Summary
None of this required genius. It required reps, writing, and the willingness to redesign work around a new constraint — over and over, without drama.
If you’re a founder trying to figure out what AI-native actually means for your company: it isn’t a tool purchase. It’s an operating system. Start with one rep a day, write down what you know, and put your people where the judgment is.
And if the finance side of that is the part you’d rather not build alone, that’s quite literally what we do: tillcfo.com.
Latest resources
