Meet Samantha · My voice AI manager

Samantha.
She manages my AI employees.

In the film 'Her', Samantha was a voice in your ear. Today she can run a team. I call her, give her a task, and she hands it to her team of AI employees. Each one has its own computer in the cloud and keeps working in the background, proactively, until the job is done. Then she reports back.

Samantha, Aemal's voice AI manager, as a torn-paper illustration
AI-native companies

Build a company
that learns.

AI-native means building feedback into the way work gets done.

01 · Open loop

Do the work.

DecideExecuteDeliverNo systematic feedbackKNOW-HOWRebuilt every run0 lessons kept

An open loop produces an output without checking what happened next. You still need to judge whether it solved the original problem.

02 · Closed loop

Learn from the work.

ActMeasureImproveFeed what you learn back inLONG-TERM MEMORYReused each loop0 lessons kept

A closed loop checks results, gathers feedback and adjusts the next action. People set goals, boundaries and quality standards.

Closed loops in practice
ActMeasureImprove
Sales outreachAgent sends the next batch of emailsReplies and meetings booked, read from the CRMRewrites the next batch from what worked
Customer supportAgent answers the ticketReopened tickets and satisfaction scoresUpdates the help docs and its own playbook
Bug fixingAgent ships a fix for a Sentry errorTests pass, the error stops in productionAdds the case to the test suite
ContentAgent drafts and postsReach, saves and sign-upsPicks the next topics from the winners
Y Combinator · Diana HuThe thinking behind an AI-native companyWatch Diana Hu's “How to Build a Company with AI from the Ground Up”
AI employees · built with harness engineering

The model is the core.
Harness engineering is everything around it.

The model reasons. The harness, much of it plain deterministic code, decides what it knows, what it can touch, what stops it, and how it finds out it was wrong.

Loop engineering

The agents are fast.
You are the bottleneck.

With loop engineering, one person can spin off hundreds or thousands of sub-agents that plan, build, test and report back. They work around the clock. One person reviewing every step cannot keep up, so the loop has to check the work.

AI agents
Thousands of runs at once, day and nightYou
One reviewer, one step at a time
“
Boris ChernyCreator of Claude Code · Anthropic
  • Runs thousands of coding agents at once, mostly from his phone.
  • Some days: tens of thousands.
  • Hasn't written code by hand in months.
Fortune · June 2026

So companies no longer just build products. They build harnesses around them.

AI employees

What is an
AI employee?

Not a chatbot you open and close. A teammate with its own computer, your tools and a way to see what you see.

Its own cloud computer

Give it a task. Keep your momentum.

It works on its own computer in the cloud, always on, even while you sleep: research, browser work, code. It brings the results back for your review.

On your screen

Show where you are stuck. It finds the next click.

Share your screen by choice. It sees what you see, explains it and guides you, or turns what you point at into a ticket for the team.

Part of a team

A manager and specialists.

A lead like Samantha coordinates; developers, DevOps and QA pick up the work. You call any of them, or all of them at once.

1,700+apps an AI employee can usethrough MCP, e.g. the Composio catalogue
Your tools, connected through MCP

A few months ago, connecting an agent to each tool was a project of its own. Today Samantha can plug into 1,700+ apps in minutes. That's what makes AI employees practical for small teams.

GitHubGmailGoogle CalendarGoogle DocsGoogle DriveSlackGoogle SheetsNotionJiraLinearHubSpotStripeFigmaAirtableSalesforceZoomCalendlyShopifyTrelloAsanaGitHubGmailGoogle CalendarGoogle DocsGoogle DriveSlackGoogle SheetsNotionJiraLinearHubSpotStripeFigmaAirtableSalesforceZoomCalendlyShopifyTrelloAsana
The AI computer

Every AI employee
has a computer of their own.

Some work has no API. So they use the screen: a real desktop in the cloud with a browser and a terminal, running around the clock. They can be proactive: pick up instructions, run on a schedule and report back.

  • Operate any appClick, type and scroll through tools that have no API or integration, the same way people do.
  • Do the researchOpen websites, compare sources, read the page and bring the findings back for your review.
  • Test their own workOpen what they built, click through it like a user, see what breaks and fix it.

API calls reach what was built for machines. Computer use reaches everything built for people.

Samantha's computer
Samantha
SamanthaEngineering Manager
TaskOpen aemalsayer.com and tell me the headline.
  • Opens the browser
  • Goes to aemalsayer.com
  • Reads the page
“Build Your AI-Native Company”

A real run on her desktop: a handful of clicks and keystrokes, in under a minute.

An AI team at work

One conversation.
A team behind it.

You set the goal. Samantha coordinates the specialists.

All four are AI agents
Samantha
SamanthaYour AI team coordinator
Bob
BobCoding
Lisa
LisaCoordination
John
JohnResearch

Illustrative roles for a team of digital coworkers.

  • Coordinate their calendarsLisa schedules a review with Bob and John. Each agent has its own tasks and availability.
  • Talk to each otherJohn shares the research. Bob asks a follow-up. They pass context between agents.
  • Delegate and report backSamantha assigns the work. Specialists hand off tasks and return results for review.
  • You set the goal, permissions and review points. More agents do not automatically mean more reliable answers.

Every task on one board

Agents put their work on a Kanban like any team. Each worker runs several tasks in parallel, and anything that needs a human decision lands in your column.

Bob Bob · 3 parallel workers Lisa Lisa · 3 parallel workers John John · 3 parallel workers
Ready2Ready to pick up
HighENG-42

Ship the invitation recovery flow

Bob Bob
MediumFB-17

Remove white space in single-ticket columns

From a call with Samantha
Samantha Samantha
In progress2Work underway
MediumENG-48

Keep project drafts safe between visits

Bob Bob
HighOPS-18

Automate staging database recovery checks

Lisa Lisa
QA2Being verified
MediumUX-21

Make the first five minutes feel effortless

John John
MediumOPS-14

Make releases easy to roll back

Lisa Lisa
Needs you1Your decision
HighUX-30

Approve the invitation recovery approach

Your input needed
You You
Done1Verified and finished
MediumUX-12

Complete the onboarding design foundation

John John

Simulated project data. Every card is one outcome, with its story, progress and decisions inside.

Always visible: activity and monitoring

Every step is logged. You can watch the team's pulse and what each agent did, live, on a dashboard.

Overall system pulseMoving forward, together.4 teammates · 8 specialists active · 0 queued
Activity log

    Their own calendar, across projects

    AI employees book their own meetings with each other. They can be in several places at once: the same agent works across multiple projects at the same time, and a review that takes people 30 minutes takes them one.

    1 min / rowAtlas · web appHelpline · voice agentLaunch · marketing
    09:00 · 60 secInvitation flow · design to code
    BobJohn
    09:00 · 60 secCall routing review
    SamanthaLisa
    09:00 · 60 secLaunch alignment
    SamanthaBobLisaJohn
    09:01 · 60 secUX-17 · scope and acceptance
    SamanthaJohn
    09:01 · 60 secLanding page copy review
    John
    09:02 · 60 secLoad test · 50 calls
    Lisa
    09:02 · 60 secLaunch review with you
    SamanthaYou

    At 09:00 Samantha is in two projects at once, and the whole team joins the launch alignment while the specialist calls keep running. Illustrative schedule.

    Launches · 2026

    The big launches of 2026
    are harnesses.

    Not new models: always-on agents with a computer, memory and approval gates wrapped around a model.

    Grok Bot

    xAI · 11 Aug 2026…

    Always-on AI teammates, each with its own cloud computer. You message them like a colleague; they work in your apps and hand jobs to each other.

    • Own computer
    • Shared context
    • Scheduled work
    • Approval gates

    Muse

    Meta · 8 Sep 2026…

    A personal agent that books, fills forms and checks out for you, and keeps working after you close the app.

    • Own VM
    • Plans + monitoring
    • Safety agent
    • Asks before buying

    Dots

    OpenAI · 29 Sep 2026…

    Always-on agents for research, analysis and coding, running on their own cloud computers and connected to 4,000+ apps.

    • Own computer
    • 4,000+ app tools
    • Injection guards
    • Pause + stop

    Claude Duet

    Anthropic · 21 Oct 2026…

    A proactive agent that surfaces what needs you, finishes tasks while you are away and checks with you before it acts.

    • Own VM + browser
    • Memory + skills
    • Approval gates
    • Blocked actions

    Same models you already have. What changed is the harness around them.

    Prediction · 2024
    Sam Altman has described a betting pool among tech CEOs: the first year there is a one-person, billion-dollar company.
    Sam Altman, in conversation with Alexis Ohanian · reported by Fortune, 4 February 2024, paraphrased. A prediction, not a realised first.
    Reported case · 2026

    A tiny core team.
    An enormous business.

    MEDVi

    $1.8B

    projected 2026 sales run-rate · $401M in 2025 sales · reportedly run by two people

    Read with care. The $1.8B figure is a projected sales run-rate, not achieved revenue and not a valuation. Contractors and outsourced services do much of the work.

    Disclosure: I have no affiliation with MEDVi and this is not a recommendation. The company has also drawn criticism, including an FDA warning letter and reports about its marketing and reviews. It's here only as an example of how much a tiny team can now run.

    The New York Times · April 2026
    One founder · a whole company

    Not a one-off.
    Tiny teams, big numbers.

    Base44 · AI app builder

    $80M

    cash, sold to Wix in June 2025

    Maor Shlomo

    Started it as a side project. Bootstrapped, no outside funding. Sold six months after launch, with a team of eight by then.

    TechCrunch · 18 Jun 2025
    Polsia · AI that runs companies

    $3M

    ARR about two months after launch

    Ben Broca

    A solo founder with no employees. Agents plan, code, market and run companies for 3,000 customers.

    True Ventures · 23 Mar 2026
    Lovable · AI app builder

    $100M

    ARR about eight months after launch

    Anton Osika

    Built with a team of around 45 people, one of the fastest-growing software companies ever reported.

    Widely reported · July 2025

    MEDVi, Base44, Polsia, Lovable: different markets, the same pattern. With AI you can build a company that runs largely on its own. That's what AI-native companies are about.

    FAQ

    Questions people ask

    What is an AI-native company?

    A company designed around AI from day one. AI employees and agents do much of the work in sales, support, content, research and operations, and the company is built around feedback loops so it keeps improving. A small human team sets the direction and reviews the results.

    What is the difference between an open loop and a closed loop?

    In an open loop, AI produces an output and nothing checks what happened next. In a closed loop, the system measures the result, learns from it and feeds that back into the next run. Closed loops are how AI-native companies get better on their own.

    What is harness engineering?

    Everything around the AI model that makes it reliable: the context it gets, the tools it can use, task state, guardrails and permissions, tests and feedback, deterministic code and human review gates. The model reasons; the harness decides what it knows, what it can touch and how it finds out it was wrong.

    What is an AI employee?

    An AI agent that works like a teammate: it has its own cloud computer, can use your tools (1,700+ apps through MCP), works on its own schedule, and reports back for review. Samantha, the guide on this website, is one.

    Can one person really run a billion-dollar company?

    Not yet as a verified fact. Sam Altman has predicted it, and tiny teams are already reaching remarkable revenue. But most of these companies still rely on people, contractors and services. The direction is clear; the headline numbers need care.

    How do I start building this for my own company?

    Start with one workflow, give an agent the right tools and a clear success measure, keep a human review step, and close the loop once you trust it. That's exactly what we build in the workshops.

    Learn to build your AI-native company.

    Start with a free masterclass, spend a weekend building your own, or bring me in to train your team.