↓ Skip to main content

Pini Shvartsman

Technology leader. AI transformation and autonomous engineering.

I see it. I build it. I own it.

Most companies need two people for this: one who sees where the work is going, and one who can build it. I do both. I drive the AI agenda at a global SaaS company, and my work is reaching every department.

Pini Shvartsman

100% of our engineering org uses the tools I build. Product, customer success and more do too.

10× the output of our best engineers. Most do 2 to 3 times the work.

80%+ of our tickets can be solved by an agent. An engineer verifies each one.

1–2 hours of an engineer's time for a ticket that took a day.

70% off the AI bill, with model routing and cost control.

My own figures, from running this work every day. Not a study.

Jobs had the vision. Woz had the soldering iron. With AI, one person can hold both.

That is the change I work on, and it is how I work. I don't predict it. I run it in production.

Vision

I see it coming.

Execution

Then I build it.

  • A change goes from idea to production in hours, with an adversarial review on the way.
  • A workplace for agents: secure cloud machines on demand, a human gate on every line, and a price on every run. The bug agent that runs on it is a team effort.
  • With it, agents can solve more than 80% of our tickets. A ticket that took a day now takes an engineer one to two hours.
  • The design is public: how my agent lanes merge and what a change must prove first.

From an empty branch to production

  1. Tuesday nightAn empty branch.
  2. 96 minutes laterThe first working milestone.
  3. Thursday64 commits in one day.
  4. SaturdayIn production.

Agents wrote the code. I set the goal, wrote the rules, read the evidence and decided what ships.

What I own
#

I drive AI enablement. I find where AI pays off and design the solution. I build the platform, roll out the tools and drive adoption across the organization.

  • The platform: where our engineers and agents work with AI.
  • The people: I stand next to the teams while their work changes, and I stay with their managers through it.
  • The rules: the AI roadmap, the standards and the governance, which I set with the CTO and the VP R&D.
  • The money: the AI budget.

Leverage
#

I build systems that multiply people.

  • Our best engineers now deliver ten times what they did. Most engineers do two to three times the work.
  • A ticket that took a day now takes an engineer one to two hours. The agent does the work. The engineer verifies it.
  • Agents write the code. I write the rules, read the evidence and answer for what ships. That is how I build a platform this fast.
  • I cut our AI bill by more than 70% with model routing and cost control.
  • I rebuilt our pipelines and cut build time in half. I also cut cloud cost by 20 to 30 percent.

AI replaces repetitive execution, not people. It gives engineers their real job back.

Leadership
#

I go first, I stand next to people, and I answer for the result.

I go first. I don’t ask a team to work in a way I have not tried myself. I built our agent platform the way I ask others to work.

I stand next to people. I train technical and non-technical teams. I sit with a team to automate their own work.

I make it stick. I ran an AI champions program and the company’s first AI hackathon. Today the work is a platform and a set of standards. The whole engineering org uses the tools. So do product, customer success and more: over 150 people. It does not depend on one team.

I answer for it. I audit my own platform, and I fix what the audit finds. When an agent breaks or the token bill jumps, it lands on me.

Adaptability
#

I change before I have to.

I took electronics apart on my bedroom floor at eight. In my teens I lived in IRC channels and wrote scripts. I have worked in tech since I was 22: IT support, NOC, freelance development, full-stack, automation, DevOps and management. I didn’t plan a career path. I just kept saying yes to whatever scared me most.

Then I joined a startup as employee #5. No infrastructure. Barely a product. I built the test framework, the CI/CD, the first deployment pipeline and the first monitoring. When we needed an offshore team, I flew to Ukraine and hired it.

That startup is now a global SaaS platform. I have changed how it ships software twice. First with pipelines. Now with agents. On the side, I have built and run my own product since 2017.

In Hebrew we call it being a Kolboynik: someone who knows a bit of everything and connects it. I am still curious about everything. Off the clock that means gadgets, science and space.

In the open
#

Most of what I know I learned by getting it wrong first. I publish it so the next team pays less for the same lesson: 80+ field notes, with no paywall and no sign-up. I also open the tools I build for my own work.

  • Product

    Invites

    Making event planning accessible to everyone. AI invitations, multilingual RSVP calls and an AI event manager, free or at a symbolic cost.

  • Open source

    MergeStory

    Turns a pull request into a short video that explains what changed, why, and what to review.

    Source code

  • Open source

    aidentity

    Keeps an admin, a test and an everyday account of the same Mac app apart, each in its own window.

    Source code

All projects

Talk to me
#

If you are in the middle of this change, talk to me. I would rather compare notes than watch you learn it the expensive way.

Invite me to your engineering team, your leadership group, your podcast or your event. I bring what is running, what broke and what it cost. I don’t bring a deck of frameworks.

Talks I can give:

  • I see it, I build it, I own it: one person, a fleet of agents.
  • Idea to production in hours: speed with proof.
  • Agents on the org chart: roles, owners and metrics.
  • Everyone trained their engineers. The gap did not move.
  • Growing the next seniors when agents write the first draft.

A short bio you can copy:

Pini Shvartsman is a technology leader focused on AI transformation and autonomous engineering. He drives AI enablement at a global SaaS company, where the whole engineering org uses the tools he builds, and so do product, customer success and more. He has changed how the company ships software twice: first with pipelines, now with agents. He shares what he learns in the open: 80+ field notes, the 14-part series Rethinking the SDLC, and open-source tools.

Write to [email protected] or message me on LinkedIn. The Let’s talk page lists what people usually bring me.

The agent writes the code. The engineer owns the outcome.

Get the next note in your feed.

Short posts on LinkedIn and X. Long notes here. Pick the place you already read.