After 8 months solo, I launched ARKLIS today: An AI CFO for small businesses
I'll just say it: building this alone for 8 months was harder than I expected, and launching it today is terrifying in the best way.
Quick background — I came from the FP&A world (financial planning & analysis). I spent years watching how much it helps a company to actually understand its numbers and plan ahead. The problem? That kind of financial intelligence has always been locked behind a full finance team, Power BI charting and high technical knowledge, which basically means only companies with cash to burn ever get it. The small businesses that need it most fly blind.
When AI got good enough, I realized the gap could finally close. So I built ARKLIS.
It connects to your accounts (QuickBooks, Stripe, Plaid, Square, payroll, etc.), watches your cash, catches things like a vendor quietly raising a recurring charge, and every morning gives you a plain-English brief on your financial health. No finance degree required.
The part I spent most of those 8 months on, and the thing I think actually makes it different: the forecasting isn't one generic model wearing different hats. A restaurant, a SaaS company, a construction firm, and a nonprofit each get a forecasting model built on the actual mechanics of their business — day-of-week swings and capacity limits for restaurants, cohort churn for SaaS, milestone billing and retainage for construction, the December giving surge for nonprofits. Ten industries, each modeled for real.
It's live today and free for early users. I'd genuinely love this community's feedback — you all have a good nose for what's real and what's fluff, and I'd rather hear it now.
Link: https://arklis.com/
Happy to answer anything about the build, the stack, the lonely-founder stuff, whatever. Ask away.
This is a strong build, especially because the industry-specific forecasting is the real differentiator.
I’d be careful not to let ARKLIS get framed as just “AI CFO for small businesses,” because that category is starting to sound broad and crowded. The sharper promise is closer to: “a daily financial operator that spots cash risk, explains what changed, and forecasts based on how your specific business actually works.”
That industry-modeling point should probably be much higher in the positioning. A restaurant owner, SaaS founder, or contractor does not care that it is AI first. They care that the system understands the mechanics of their business better than a generic dashboard.
For early users, I’d probably pick one vertical first instead of pushing all ten. Restaurants or small SaaS could be a cleaner wedge, because the pain is easy to explain and the value shows up quickly.