Assaf Ronen

I'm the President of the HelloFresh Group and sit on our management board. Over the last 6 months we've quietly rebuilt how we operate with AI. Not a pilot. Not a slide in a board deck. Real P&L impact. Here's what's working, what broke, and what nobody warns you about 🧵

@ju1lzzz True far beyond design. In every function I run, the divide isn't senior vs. junior — it's the people who adopted AI vs. the ones waiting for permission. The tool doesn't threaten your job; refusing to pick it up does.

AI isn't replacing the middle designer. It's replacing the one who refuses to operate it The video's right that the €2,000 tier gets eaten first. It's wrong about where that work goes. That budget doesn't vanish. The owner behind it was never going to learn Framer AI or write clean prompts - they don't have the time, and they never wanted a design, they wanted the task done. So the tier stays open, just for a different role: the operator who runs the same skills I use to move 3-4x faster and ships the whole thing turnkey in an afternoon. "Go premium" is only half the advice. The half nobody says: the bottom doesn't close, it just stops paying designers and starts paying operators. So the real split isn't top segment vs bottom. It's who runs the AI versus who competes with it. Which side are you building toward?

@businessbarista As one of those execs — the most useful question isn't "what can AI do for us?" It's "where in our operation is the work repetitive, rules-heavy, and high-volume?" Point AI there first. The flashy use cases are rarely the highest ROI.

My team spends all day talking AI with enterprise execs. I asked them to share the most common questions they get. Here's what we're hearing from the field: • How to properly build a UAT suite that can test not only the new software we are building but also account for the AI features in some benchmark eval suite? • How can I provide guardrails and direction to enable the front line to develop useful tools and applications for the business? • How do you deal with fragmented data? What is your process to unify without a massive overhaul of our orgs data architecture? • How do I stand up the internal motion to drive AI tools and workflows when everyone also has their day job? • How do we know an agent that we create is actually "good"? How do we measure that? How can we improve it? • How can I give employees full agency to create without impacting mission-critical systems/workflows/etc.? • How do I uniformly transform a multi-thousand person org to adopt ai? How do we not leave anyone behind? • How do I ensure that I roll out claude code and cowork securely without putting my company data at risk? • How do I control spend of ai usage across my organization without limiting my employees' productivity? • What does the operating model have to look like with my direct reports as well as the org with AI? • How do I start controlling token spend and how do I think about attributing value to a token? • How do we develop a central company brain to capture embedded organizational tacit knowledge? • What are the best ways to be multi-model and have a multi threaded approach to partnerships? • How can non technical people access, change, and iterate on apps they did not build? • When an agent does eight hours of work, how does a human check it in eight minutes? • What are the big investments I need to make in my data to make AI effective? • How do I protect my data while still having the harness of cowork and code? • How do employees in different business units edit, manage their own skills? • How do we adopt AI so that we aren't vendor locked with one frontier lab? • How do we ensure our AI usage is safe (infra & security controls)? • What data is safe to put in (especially sensitive functions)? • Whats the path from AI Literate, to AI Enabled, to AI First? • How do we get people excited vs scared to lose their jobs? • How can AI apply when I'm in a highly regulated industry? • As a CEO, what do I need to know about AI to run my org? • Should I hire a team vs. work with an external partner? • What's the bleeding edge of applied AI look like today? • How do we protect our proprietary data when using ai? • How do we manage costs, and prevent runaway sessions? • When processes are the problem, where do I start? • How do we let people access internal data safely? • Should I allow Skill creation? Artifact creation? • Who should I give access to Claude Code or Codex? • What does the cutting edge of AI SDLC look like? • How do I "sell" AI internally within my company? • How do I make big bets and also avoid lock in? • How do I know which models are actually good? • How to build model agnostic capabilities? • Should I be implementing spending limits? • Who owns a build after it's deployed? • How to capture the full scope of ROI? • How do we prioritize use-cases? • What are other companies doing? • Who should own AI internally? • How do we distribute skills? • What is an agent harness? • How do we govern AI? • Are we behind?

@jasonlk The build is the easy part now. The real unlock was how fast the team adapted around the agents — the org rewiring is harder than the tech, and it's where most companies stall. Fun 12 months on our side too.

Few things have been more fun in my career in software and tech than the past 12 months We've gone from 1 basic agent to 21+ AI Agents running all of SaaStr, much of SaaStr Fund, and so much more But the cognitive load is at the edge. It's at the limit. I don't think we can manage one more independent agent. And everything now is 24x7. The backlog. The agents never stop working. Mornings, nights, it's all agents, all the time. It's great, it's so cool, and I hate the term "burn out." That called life when you are building. But it's a level of constant cognitive load I've never really experienced before.

@alliekmiller From inside a company actually doing this: yes, teams are getting smaller — but the ones that thrive aren't cutting to cut. They're concentrating people on judgment, taste, and the calls a model can't make. The rote work leaves; the hard thinking stays.

Everyone has a theory on how AI will impact the job market. Here's what I'm seeing: 1/ Everywhere I look, teams are getting smaller. That does not mean overall headcount is reducing or unemployment is increasing, but my anecdotal experience is telling me that the smallest meaningful operating unit is shrinking by a factor of 2 or more. 2/ Some government groups I have spoken with (not just in the USA) are preparing for a surge in unemployment benefit claims. That does not mean it's happening, nor does it mean their plans are sound, but it's helpful to know that is one outcome they're considering. 3/ Everyone has incentives. The AI labs, the CEOs, the researchers, the banks, the government, heads of state, even you reading this. It's important to keep that in mind, especially as tunes change (ex: from mass job loss to mass job gains). 4/ Even if AI creates new jobs, it is likely that they will require new skills or be in new areas of expertise. That means there will still be a skills gap. That skills gap feels like it's growing wider every day. More jobs existing is not the same as high employment aka it's not just about jobs existing, it's whether they get filled (and yes, I wrote that and not AI). Outside of a few hundred content creators and a couple of enterprises, I'm not seeing enough movement to close the gap. 5/ Keep pushing enterprise leaders to provide a timeline on their job predictions. They'll refuse at first because they are PR-trained. Push back. Ask for a range, ask if it's over or under 10 years, ask what would have to be true for it to happen in 3 years versus 10 versus 50.

The "company with zero employees" dream is the wrong dream. The real shift isn't subtraction — it's leverage. A team of 10 now moves like a team of 100. The winners won't brag about zero headcount. They'll have small teams that are suddenly dangerous.

So we stopped running one big "transformation program." Instead: rolling thunder — small AI wins shipped weekly. Momentum compounds faster than any roadmap. I'll share them here as they land — numbers and scars included. Follow if you want the operator's view, not the hype.

What surprised me most: The bottleneck was never the model. It's your data, your approval chains, and your org's tolerance for "good enough, shipped today" over "perfect, shipped in Q3." The tech is ready before the company is.

I'm the President of the HelloFresh Group and sit on our management board. Over the last 6 months we've quietly rebuilt how we operate with AI. Not a pilot. Not a slide in a board deck. Real P&L impact. Here's what's working, what broke, and what nobody warns you about 🧵

I sit on two boards. Same question in both rooms lately, and it's usually the wrong one: “do we have an AI policy?” The real question: “can we defend this AI-influenced decision in 12 months?” Policy is paperwork. Defensibility is governance.

Last week our teams demoed AI agents they built themselves — a security agent, supply-chain automation. No mandate. Just people building. We're a tech-food company. This is what that looks like on a random Tuesday, not a press release.

@satyanadella This is the number people should be reacting to, not the demo reels. 60%+ in production changes the conversation from “is AI ready” to “why isn't yours in production yet.” The gap between pilot and this is where most companies get caught flat-footed.

@reachvaldo Art

You don’t have The Art of AI Funnels? like + reply “Art” and I’ll DM it to you for FREE.

@mikefutia Banana

Nano Banana + Veo 3 + n8n is legitimately wild 🤯 This automation generates hundreds of UGC videos from a single product photo. Fully automated inside n8n. Perfect for DTC brands & creative agencies scaling paid social. Most brands face the same bottleneck: You need 50+ creative variations to beat ad fatigue. But creating scripts, filming, and editing that volume takes weeks and costs thousands. This n8n system solves it: → Drop ONE product image into n8n form → Nano Banana creates 20-50 visual variations automatically → Each variation flows into Veo 3 for video generation → AI generates natural UGC scenarios (unboxings, demos, testimonials) → Videos auto-saved to Box for instant access No scripting. No filming. No editing bottlenecks. What you get: Dozens of UGC videos from one upload → Costs pennies per video → Full commercial ownership → Perfect for testing angles at scale Built 100% in n8n. Want the complete workflow? > Comment "BANANA" > Like this post And I'll send it over (must be following so I can DM)

Big news: Our Meals with Meaning program just hit 10 million meals served! This was a massive team effort from our partners, volunteers, and employees. Together, we're making a real impact in the fight against hunger. #HungerActionMonth youtu.be/awHbJJ1RU1E

Hello fresh The most customer centric FoodTech company in the world investing an unprecedented amount in increasing potions , variety , health and taste to make sure that every dinner with hello fresh will leave all family members amazed.

Bigger. Healthier. Tastier. Introducing the new HelloFresh bwnews.pr/47fRN3T

Thrilled to see @SoFi included in this year's @TIME list of the 100 Most Influential Companies of 2022. We are on a mission to help people get their money right and we're only just getting started...‼️ #TIME100Companies

🔥🔥👇🔥🔥