I thought I knew how to manage a team. Then I had to do it remotely with AI engineers. Here are the 10 hard-won lessons that saved my sanity and my company.
I’ve been a serial entrepreneur in Silicon Valley for over a decade. I’ve had two successful exits: RemoteTeam, which was acquired by Gusto, and MovieLaLa, acquired by Gfycat. I’ve also been an angel investor in over 200 companies, including some of the biggest names in AI like Anthropic, OpenAI, Scale AI, and Hugging Face. But nothing prepared me for the challenge of managing a remote AI team.
It’s a different beast altogether. The hype around AI is deafening, but the reality of building and managing a team of AI experts is a messy, human endeavor. After years of trial and error, here are the 10 most surprising and effective lessons I’ve learned.
1. Hire for deep expertise, not just “AI” hype.
In the early days, I made the mistake of hiring people who were great at talking about AI but couldn’t deliver. They knew all the buzzwords, but when it came to actually building something, they were lost. I remember one candidate who aced the interview, talking a big game about transformers and diffusion models. He had a PhD from a top university and a string of impressive-sounding publications. But when we put him on a real-world project, he couldn’t even get a simple model to train. He was a master of theory but a novice in practice.
Now, I hire for deep, narrow expertise. I’d rather have an engineer who knows everything there is to know about a specific type of model than a “generalist” who knows a little bit about everything. The AI generalist is a myth. You need specialists. For example, we recently hired an engineer who had spent her entire career working on natural language processing for low-resource languages. She wasn’t a big name in the AI world, but her deep expertise was exactly what we needed for a new project. She was able to build a model that outperformed the state-of-the-art in just a few weeks.
2. Over-communicate the “why,” not just the “what.”
AI engineers are not code monkeys. They’re problem solvers. They’re motivated by tackling hard, interesting challenges. If you just give them a list of tasks to complete, they’ll get bored and disengaged. You need to explain the “why” behind the work. What’s the business context? What problem are we trying to solve for our customers? Who are we trying to help?
I once had a project go completely off the rails because the team didn’t understand the business context. We were building a model to predict customer churn. The team built a technically brilliant model that was 99% accurate. But it was completely useless. Why? Because it was a black box. We couldn’t explain to our customers why they were at risk of churning. We had to scrap the whole project and start over. It was a painful lesson, but it taught me the importance of over-communicating the “why.” Now, we start every project with a detailed brief that outlines the business context, the customer problem, and the desired outcome.
3. Your best engineers are probably the quietest.
In an office, it’s easy to see who’s working hard and who’s not. But in a remote team, it’s much harder to gauge performance. The loudest person in the room is not always the most productive. In fact, I’ve found that my best engineers are often the quietest. They’re the ones who are heads-down, focused on their work, and not interested in self-promotion.
I almost overlooked one of my best engineers because he was so quiet. He rarely spoke up in meetings, and his updates were always brief and to the point. But when I looked at his code, I was blown away. He was a true 10x engineer. He was working on a complex new feature that nobody else on the team could figure out. He had been quietly chipping away at it for weeks, and he had finally cracked it. I learned that I needed to create systems to hear from everyone, not just the loudest voices. Now, we use a combination of one-on-one check-ins, written updates, and anonymous surveys to make sure that everyone has a chance to be heard.
4. Ditch the daily stand-ups.
I used to be a big believer in daily stand-ups. I thought they were essential for keeping the team aligned and on track. But in a remote team, they quickly became a waste of time. People would either ramble on for too long or give vague, unhelpful updates. It was a ritual that had lost its meaning. The time zone differences also made it a nightmare to schedule. Someone was always having to join at an inconvenient time.
So, we ditched the daily stand-ups and replaced them with something better: asynchronous updates in Slack and a weekly deep-dive session. The async updates allow people to share their progress on their own time, and the weekly deep-dive is a chance for us to tackle hard problems together. It’s a much more effective system for a remote team. The async updates are more detailed and thoughtful than the verbal updates we used to get in stand-ups, and the weekly deep-dive is a much more productive use of our time together.
5. Time zones are a feature, not a bug.
When I first started building a remote team, I was obsessed with time zones. I tried to hire people who were all in the same time zone, or at least within a few hours of each other. I thought it would be impossible to collaborate effectively with a team that was spread out across the globe.
I was wrong. I’ve learned that time zones can be a huge advantage if you embrace them. We now have a team that’s spread out across 10 different time zones, from California to Turkey. This allows us to have a 24-hour development cycle. When the team in San Francisco is signing off, the team in Istanbul is just getting started. We use a combination of tools like Slack, Asana, and GitHub to hand off work between teams. It’s a powerful way to accelerate development. We can literally get a full day’s work done while we sleep.
6. Invest in writing skills.
In a remote team, writing is thinking. It’s the primary way that we communicate with each other. A poorly written spec can lead to a major setback. A well-written document can align the entire team and unlock a new level of productivity.
I’ve learned to invest in writing skills. I encourage my team to write, and I give them feedback on their writing. I also make sure that we have a culture of documentation. We document everything, from our code to our processes. We have a company-wide wiki where we share our knowledge. It’s a lot of work, but it’s worth it. A well-written document is a force multiplier. It can be read by anyone, at any time, and it can be referenced over and over again. It’s the most scalable way to communicate.
7. Don’t just measure output, measure progress.
It’s easy to measure output. You can track lines of code, number of commits, or number of features shipped. But output is a vanity metric. It doesn’t tell you anything about whether you’re actually making progress towards your goals.
I’ve learned to focus on measuring progress, not output. We set clear goals and milestones, and we track our progress against them. This allows us to see if we’re on track, and it helps us to identify and address problems early on. It’s a much more meaningful way to measure performance. For example, instead of tracking the number of models we train, we track the accuracy of our models on a held-out test set. This tells us whether we’re actually making progress on our goal of building more accurate models.
8. Your team’s mental health is your #1 priority.
Remote work can be isolating. It’s easy for people to feel disconnected and burned out. As a manager, it’s your job to be proactive about supporting your team’s mental health. This means creating a culture of psychological safety, where people feel comfortable talking about their struggles. It also means encouraging people to take time off and to set boundaries between work and life.
I once had a team member who was on the verge of burnout. He was working long hours, and he was clearly struggling. I had to intervene and force him to take a vacation. It was a wake-up call for me. I realized that I needed to be more proactive about supporting my team’s mental health. Now, we have a number of initiatives in place to support our team’s mental health. We offer a flexible work schedule, a generous vacation policy, and access to mental health resources. We also have regular one-on-ones where we check in with each other on a personal level.
9. Celebrate the small wins.
In a remote team, it’s easy to lose momentum. You don’t have the same opportunities for spontaneous celebration that you have in an office. You have to be intentional about celebrating the small wins. This could be anything from a successful model training to a positive customer review.
We have a #wins channel in Slack where people can share their accomplishments. We also make sure to celebrate our wins in our weekly all-hands meeting. It’s a small thing, but it makes a big difference in morale. It helps to create a sense of shared purpose and to remind us that we’re all in this together. We also have a tradition of sending a small gift to team members who have gone above and beyond. It’s a small gesture, but it’s a powerful way to show our appreciation.
10. The best tool is the one your team actually uses.
I’m a self-proclaimed productivity nerd. I love trying out new tools and apps. But I’ve learned that the best tool is the one that your team actually uses. It doesn’t matter how fancy or powerful a tool is if nobody on your team is using it.
I once tried to implement a fancy new project management tool that had all the bells and whistles. It had Gantt charts, resource allocation, and all sorts of other features that I thought would be great for our team. But nobody on my team used it. They preferred to use a simple Trello board. I had to swallow my pride and admit that the simpler tool was the better tool for us. Don’t get seduced by the latest and greatest tools. Focus on what works for your team. The best tool is the one that helps your team to get their work done, not the one that has the most features.
The Future is Remote and AI-Powered
Managing a remote AI team is not for the faint of heart. It’s a constant challenge, but it’s also incredibly rewarding. The future of work is remote, and the future of technology is AI. If you can figure out how to build and manage a successful remote AI team, you’ll be unstoppable. The lessons I've shared are not just about managing a team of AI engineers. They're about leading a team of creative, passionate problem-solvers in a remote world. They're about building a culture of trust, transparency, and collaboration. And they're about embracing the future of work, not fighting it.
Frequently Asked Questions
Which item on this list has the highest impact?
It depends on your stage and context, but in my experience, the items near the top of the list tend to have the broadest applicability. That said, sometimes the less obvious items create the biggest breakthroughs for specific situations.
How were these items selected?
Each item on this list comes from direct experience, either from building my own companies or from patterns I've observed across the 200+ startups I've invested in. I prioritize practical, actionable items over theoretical concepts.
Are these recommendations still relevant in 2026?
Absolutely. While specific tools and tactics change, the underlying principles remain consistent. I update my thinking regularly based on what I'm seeing in the market and across my portfolio companies.