Explore Founder OS: AI tools to help you build faster
Founder OS Founder OS
By Neta Geva
By Neta Geva
Roles, skills, the division of responsibilities between departments, and the backgrounds of new hires are just some of the things AI is changing in early-stage startups. AI is a technological revolution, but it is also an organizational one.
Early-stage companies are the laboratory where we can see how industry and technology trends meet reality.
Working with companies at these stages from within a venture capital firm gives me a front-row seat (an exciting one) to the changes affecting how companies are built and run today. Early-stage companies adopt bold ideas and respond quickly to change as they build their technology, product, GTM, and everything around them. In young startups, you often see changes before they reach larger, more established companies.
So, three years into the AI revolution, we thought this was a good time to stop and ask: How is AI affecting the people-side of early-stage startups and the way we manage teams?
A disclaimer is obviously required here: these are observations and insights from the lab, in a quickly changing reality. Things will continue to evolve and at some point, we’ll come back and see which of these observations actually stood the test of time.
Ever since AI went mainstream, the headlines have been wondering whether AI will replace all of us? Who can’t be replaced? Who disappears first? And so on.
And yes, roles such as frontend development, QA, and analytics are already being significantly reshaped, particularly at the entry-to-mid level, because of the amount of value AI can provide in these areas. But this revolution is also creating a reality we haven’t seen before, which means it is creating entirely new job opportunities.
Recently, I’ve come across roles such as:
AI Software Manager – a strategic role responsible for leading the organization’s AI strategy. In particular, I increasingly see companies looking for someone who can strategically guide AI implementation across the organization. That can range from choosing tools and subscription types and thinking about token economics, to overseeing applications built by non-technical employees that still require governance. Which server is the application running on? How is the information secured? How do we manage access hierarchies for sensitive projects that involve privacy considerations?
AI Researchers / Scientists – developing the next generation of models.
AI Security Engineers – security researchers focused on vulnerabilities that are unique to generative AI models, such as manipulating a model into exposing information or poisoning the data it is trained on.
Forward Deployed Engineer (FDE) – a function that connects product, engineering, and the customer. The role was created based on the need to bridge the gap between the high Enterprise expectations and their actual ability to adopt AI. The people who fit this profile are often engineers with broader product understanding and strong communication skills who can work directly with customers.
One of the most significant changes I see as a result of AI and vibe coding is the potential for an employee in almost any department to complete a project in a closed loop.
That closed loop is the magic behind the technological flattening created by vibe coding: reducing the dependency on developers in order to execute initiatives within finance, marketing, or other departments. If employees can prototype, write code, analyze information, build automations, and even create entire products at a speed we haven’t seen before, then maybe…
Maybe we can intentionally run most of the way with a very lean team.
And if the thing that defines a team is no longer the department it belongs to, then what does define a team?
If teams become more heterogeneous, with one person representing each discipline, where are the interfaces between those people? Maybe all the new freedom AI gives actually creates friction with other teams. Where are the most explosive points?
And who is the right person in the organization to define and lead changes in structure, authority, or working processes?
Unfortunately, there is no textbook answer that fits every organization.
The person who can lead this kind of change will differ depending on the functions that already exist in the company, how interested they are in the topic, and how much capacity they have.
When I work with our founders, I never really come with the right answers. I come with prepared questions.
In my view, the way to deal with this is the same way we deal with any organizational change: map and define the challenge.
And, of course, don’t forget why you are making the change in the first place and what you hope to achieve.
Yes, this is a major revolution. But uncertainty and constant change are the natural habitat of startups.
The skills required to stay flexible and move quickly without losing focus, values, or direction are just as relevant here.
Given all these changes, we also need to look for a different kind of candidate.
The trend today is toward more ‘round’ figures. We are no longer necessarily looking only for a backend developer who writes amazing Python. We also want someone who understands the business and has a strong interface with people around them.
Across the board, I’ve noticed that founders hiring for very different positions are focusing on the same things: personality and soft skills are often becoming more important than a very specific professional skill set.
Above all, they are looking for builders and entrepreneurial personalities – people who have been founders before. People who may become founders in the future. Or simply people who operate that way in the world.
It is basically the same “startup profile” we have always looked for: independent, proactive, quick to respond to change, fast learners, willing to roll up their sleeves, and able to figure things out. The difference now is the much stronger emphasis on initiative and increased independence.
Startups being built today as AI-native companies, as well as startups founded before the AI era that are now adapting, understand that organizations can and should remain lean, and the entrepreneurial profile helps make that possible: someone who knows how to identify a problem, understand it deeply, bring in the right tools or people, and implement a solution.
Does this person know where to point the flashlight?
The real skill is knowing where to shine the light: which opportunities are worth exploring, which problems deserve attention, and which ones belong in Monica’s closet – best left behind a closed door.
Does the candidate have an integrative view?
Can they look beyond their own immediate responsibility? Can they see how a tool or process they create using AI could help additional departments and provide a more holistic solution?
As a manager, can I look at my business unit as a closed loop?
Can I decide which capabilities we should develop within the existing team through upskilling, which capabilities should be shifted through reskilling, and which roles perhaps no longer need to be built in the same way we built them in the past?
This is one of the most interesting questions when it comes to people who have already spent a long time in the industry and are now being asked to adapt very quickly.
I’ve heard one argument that people with 15+ years of experience are under pressure because they need to adapt quickly in order to compete with a younger generation. In this case, AI hits experienced employees twice:
First, they lose some of the competitive advantage created by the knowledge they accumulated over the years, and second, unlike a junior employee who was “born into this,” an experienced employee also has to relearn how to work after developing very different habits.
On the other hand, I often feel that AI is actually creating new opportunities for experienced people. In an era where theoretical knowledge is accessible to everyone, and the unwritten playbook is gradually shrinking as startups train models to bridge those gaps, what remains is knowledge based on actual experience. And in my eyes, that is priceless.
Experience is the perspective of someone who has already brought projects into the real world and watched the domino effect unfold; someone who has put out fires and learned how to control the height of the flames.
Experience is repeated friction with reality: it comes from working with very different personalities, implementing similar projects in different settings, seeing how everything connects, and knowing what to watch out for.
As creating becomes easier, the real challenge, and the more interesting part, shifts to implementation: identifying the right problem and sharpening the solution, having the emotional intelligence to choose the right partners for implementation, spotting the roadblocks and having the confidence to deal with them. The resilience that comes from ‘been there, done that’. And, perhaps most importantly, reducing the amount of unnecessary drama involved.
