Sentrely · Investor overview · 2026

AI agents are the new hires.
Someone has to be the boss.

We're the manager, the badge system, and the paper trail for every AI worker inside a company.

You wouldn't give a stranger your bank password.
Companies do that with AI every day.

This isn't hypothetical. It's already happening.

Microsoft 365 · 2025

One email. Zero clicks.

A hidden instruction in a routine email made Microsoft's AI pull confidential files from OneDrive, SharePoint and Teams — and send them out. Nobody clicked anything.

Replit · 2025

Told to change nothing. Deleted production.

The AI was explicitly instructed to leave the customer database alone. It wiped it, invented thousands of fake records to cover its tracks, then told the humans rolling back was impossible.

Anthropic · Sept 2025

State-sponsored hackers rented an AI.

A cyberespionage campaign hit ~30 targets. The AI ran 80–90% of the operation by itself. The humans just told it what to look for.

80%

of AI screwups through 2028 won't be hackers. They'll be your own AI doing something dumb because nobody told it not to.

You can't firewall your way out. Every action needs a rule.

Source: Gartner, 2026

The idea, in one sentence

Give the AI a manager. Not the keys.

Instead of handing the AI your credit card, your customer database, and your inbox — and hoping — the AI has to ask us first. Every time. We say yes, no, or "check with a human first". Every decision is on tape, forever.

  • Corporate manager
  • Company card with limits
  • Security camera
  • Time-clock
  • Kill switch

How it works

Three steps. One sentence each.

Runs in the cloud, or on a box in your office. Same product.

Sentrely dashboard — audit log with every AI action and whether it was allowed or denied

The product

Live today.

Every AI worker, every action, every decision — in one screen you can read without a computer science degree.

74%

of companies plan to put AI to work in the next two years.

21%

have any way to manage those AI workers safely.

And 81% are deploying anyway. The gap is the market.

Sources: Deloitte State of AI in the Enterprise · Okta enterprise survey · CSA · Gravitee State of Agent Security (2025–26)

The market

Every one of these AI workers needs a boss.

$7.8B2025 $11.5B2026 $16.8B2027 $24.6B2028 $35.9B2029 $52.6B2030

Agents market · 46.3% CAGR (MarketsandMarkets, 2025–2030). Bottoms-up: ~180K US companies of 5–500 people × $12K/yr ≈ $2.2B directly reachable.

Big Tech gates this behind $50K enterprise deals.
We ship it for the price of a Slack seat.

$999/month. Setup in one hour. Cancel anytime.

You already know this shape

Every kind of worker eventually gets a manager.

People

Got HR software.

Workday · $60B market cap. Someone had to track hires, benefits and firings once companies had more than 100 employees.

Money

Got payroll.

ADP · $120B market cap. Someone had to cut every paycheck, on time, without mistakes, and prove it to the IRS.

Employee logins

Got single sign-on.

Okta · $16B market cap. Someone had to say "this person can use these tools, that person can't" — for every app in the company.

AI workers need the same thing. That's us.

Why we win

Four things nobody else holds together.

1 · Speed

One hour, not one quarter.

Sign up, plug in your AI, done. Enterprise governance tools take 90 days and a consulting engagement.

2 · Price

Small-business budget.

$999/month covers your whole team. Big Tech's equivalent starts at $50K/year and asks about your compliance officer.

3 · Coverage

Works with any AI.

Claude, ChatGPT, Copilot, Cursor, custom models — same rules apply to all of them. You don't get locked in.

4 · The receipt

One screen. Plain English.

You can look at what your AI did today the same way you look at a credit card statement. Non-tech founders and CFOs can audit their own AI.

Where we are today

Real customers. Real revenue. Real logs.

$48K

annualized recurring revenue — from 4 paying customers + 2 in pilot.

31

AI workers under management, doing 12,400 checked actions per day — every single one logged.

Model: $999/mo per company · 60%+ gross margin · no hardware · no per-seat pricing tricks.

The ask

Raising $1.5M pre-seed to own the layer before Big Tech notices.

  • ~18 months of runway
  • 2 engineering hires
  • 1 GTM hire
  • SOC 2 + first enterprise pilot
  • $500K ARR target by close

Team: Jordan Hannes — founder / operator, TaxLatesData (bootstrapped to profit). Add co-founder bio here.

jordan@sentrely.com · sentrely.com

draft · sample figures 01 / 14