Adi Chen

Building | Bootstrapped cybersecurity company to hundreds of clients

[Adi Chen @_adichen]

The Future of Work:
What If Every Knowledge Worker Could Scale Their Expertise Beyond Their Working Hours?

If you’re wondering how to leverage AI as an individual, start by thinking about Agent as a Service.

Find something that you are really good at, encode it to an agent that would drive an outcome.

Instead of trading your time, you’re selling your expertise as software.

We’ve already run this experiment in Sales, GTM and Engineering.

We’re looking to partner with exceptional knowledge workers.

If you’re an expert with a repeatable way of creating value, let’s turn your expertise into an AI employee that companies can hire.

If that sounds interesting, our DMs are open.

[Adi Chen @_adichen]

Proof that America is the most efficient market in the world:

Yesterday, I spoke with a 67-year-old CEO running a company that generates more than $50M in annual revenue.

He told me he doesn’t know where to start with AI, but he knows he must.

Don’t be the company that wakes up one day and realizes its operating model has become too expensive to compete

[Adi Chen @_adichen]

The real value of AI will emerge at the department level, not through individual or task-level productivity.
It will make management executable, enable massive scale, and reduce operational overhead. Once executives understand this shift, AI will become a true engine for business growth.

[Adi Chen @_adichen]

Now I’m building our sales workflow.

Instead of giving every salesperson an AI assistant, I’m redesigning the interface through which sales happens in our company.

I defined how the ideal salesperson should operate, encoded that workflow into agents, and made it the default way our team works:

1) Every new lead receives an immediate, personalized response.
2) Before every booked call, the salesperson receives a pitch deck tailored to the prospect and updated with our latest company and sales data.
3) Every call teaches the entire sales team. If someone discovers a new competitive angle, objection, or message that resonates, it becomes available to everyone.

As a CEO, this gives me something much more valuable than productivity:

1) Knowledge no longer disappears inside individual employees.
2) My management becomes executable. Instead of repeatedly telling the team how I want sales to be done, I encode those standards into the workflow.
3) Onboarding becomes much faster. Every new hire immediately gains access to the department’s accumulated knowledge and can learn from the best sellers.
4) When someone leaves, I no longer lose valuable knowledge I didn’t even know they had.

The real shift isn’t giving every employee access to AI. It’s turning the way the company should operate into a system that learns and improves every time the team works.

If you’re building a startup and designing your company from scratch, I’d love to share what we’re building and build it with you.

[Harry Stebbings @HarryStebbings]

"We replaced Salesforce with a vibe-coded CRM built for our own workflows.

The custom system integrated our AI agents more effectively, worked better for the team, and made Salesforce unnecessary.

That decision cut a $600,000 annual software bill to zero."

Is this an anomaly or the start of a much larger trend @chamath @Avishai_ab @jasonlk @benioff

[Adi Chen @_adichen]

Is it the name of OpenAI’s new revolutionary hardware device or the git branch they are working on?

[OpenAI Developers @OpenAIDevs]

Meet kbd-1.0-codex-micro, built with @work_louder.

Map the buttons and joystick to your workflow, and keep your pinned chats in view.

Get yours before stock returns 410.

[Adi Chen @_adichen]

OpenAI decided to take almost every bit of control you have to teach you one thing -
the agent is in control.

Back in 1983 Microsoft taught the world how to use their new operating system using a new hardware device called the mouse, and it changed the world.

Today, OpenAI is teaching the world how to use AI agents by introducing their new hardware.

Will we remember this moment 40 years from now?

[OpenAI Developers @OpenAIDevs]

Meet kbd-1.0-codex-micro, built with @work_louder.

Map the buttons and joystick to your workflow, and keep your pinned chats in view.

Get yours before stock returns 410.

[Adi Chen @_adichen]

What I love about being a founder today is that it has never been easier to encode quality into how a company operates.
For every important process, I can define how it should work, turn that logic into an agent, and give it to the team.

People use it, improve it, and the process gets better without the original knowledge being lost.

One of the hardest parts of scaling a company is that information degrades as the organization grows.
Context gets trapped in people’s heads. Standards become inconsistent, and you lose control.

[Adi Chen @_adichen]

Very little value will be captured if we treat agents as individual productivity tools.

Giving every employee their own agent only makes fragmented work happen faster. Everyone still works differently, and the company still needs managers, processes, and coordination to keep everything aligned.

The real transformative potential of agents is at the company level: giving leadership the ability to code how the organization should operate its processes, standards, and definition of quality.

Once executives understand that, they will start capturing the real value of agents.

[Adi Chen @_adichen]

The biggest opportunity in agents may be letting CEOs program how their companies operate.

Today I spoke with a founder who wants to make support, sales, and operations more agentic.

Three things stood out:
1) No one wants to own the transformation.
Engineering wants to build the product. Sales wants to sell. Support wants to serve customers.
So the company needs to hire someone or bring in an external partner. There is already a budget for ownership

2) Personal agents create fragmentation
When every employee builds their own workflows, the company simply recreates the same inconsistency it already has with people different processes, standards, and levels of quality.

3) Shared agents make management executable.
A CEO can encode how leads should be researched, how customers should be supported, what quality looks like.
A CEO can probably, for the first time, truly codify and enforce how his company operates.
Instead of every employee inventing their own workflow, the company gives everyone the same operating standards.

[Adi Chen @_adichen]

I really hope AI kills coding, because I completely forgot how to code

[Adi Chen @_adichen]

יש לי תחושה שתמיד המודלים יהיו גרועים במרקטינג. המודלים לומדים מהממוצע, ומותגים גדולים נבנים מלברוח מהממוצע, בלהגיד משהו חדש שאף אחד לא אמר כי אם כולם יגידו את אותו הדבר הכל יהיה interchangeable

תבנו identity למוצר שלכם, לעצמכם
אל תהיו עוד founder slop

[Adi Chen @_adichen]

מסכימה. השוני היום הוא שהעלות למחקר פרסונה ודומיין ירדה משמעותית. פעם הגישה לידע הייתה צוואר הבקבוק
(שעות של שיחות ובניית אמון עם לקוחות, רק כדי להבין את הבעיות שלהם)

היום אפשר להגיע להבנה ראשונית טובה של הדומיין בכמה שעות עם קלוד, ואז להגיע לשיחות עם לקוחות הרבה יותר מוכנים. זה מאפשר איטרציה מהירה יותר, שיחה ראשונית הרבה יותר מפוקסת, ויוצר תחרות גם באזורים שפעם נשלטו בעיקר על ידי אמון וגישה

[Adi Chen @_adichen]

היום נפגשתי עם לקוח לגבי מוצר שבניתי עם AI שיכול להתאים לתחום שלו, השאלה הראשונה שהוא שאל אותי הייתה אם עבדתי בתחום בעבר
עניתי שלא.
הוא היה ממש מופתע מרמת הידע שלי על התחום שלו, מההבנה של איך הדברים עובדים הבעיות והפתרונות שהבאתי.
זה גרם לי לחשוב שאולי היום צריך פחות Design Partners מבעבר. אם אתה מגיע מוכן מספיק לפגישה הראשונה, עם מחקר מעמיק והבנה טובה של הדומיין, פתאום אתה מוצא את עצמך בשיחת מכירה במקום בשיחת ולידציה

[Adi Chen @_adichen]

@aloncarmel עוד מקום מעניין הוא דאטה. אינטגרציות נהיות מהר יחסית commodity, אבל דאטה איכותי שנצבר לאורך זמן וקשה לשכפל הוא יתרון תחרותי אמיתי
או דאטה שעל מנת להשיג אותו (והרבה ממנו!) לוקח המון זמן