I built the first version of my grandfather’s AI clone just to ask for his advice. He passed away before I knew how to ask the right questions. This time, I wanted answers. That prototype sent me down the rabbit hole. Now anyone can do the same.

Trying to get to know yourself? Don’t meditate. Don’t journal. Start a company. Building something from zero forces you to meet every version of yourself—the good, the bad, and the delusional.

I couldn’t ask my grandfather for advice. He had a stroke and I never learned Farsi. So I built his AI clone and asked him anyway. That one experiment changed everything. That’s where Delphi started.

Deli isn’t an AI chatbot. It’s a new layer of identity. One where your expertise can scale, teach, and even earn — while you sleep. We’re not building AI entertainers. We’re building AI extensions of real minds.

Clones that sound human. Clones that hold you accountable. Clones that match you to mentors like Spotify recommends songs. That’s the future we’re building at Delphi.

I work 7 days a week. Not for grind culture. Not for optics. But because I’d regret doing anything else.

When the mission matters, your fuel changes. It’s not energy, it’s belief.

Keith taught me to think in power laws. Most advice comes from people optimizing for the average, not the outliers. The real signal is in the weird, obsessive inputs no one wants to do. That’s what I look for in founders. That’s what gets funded.

Delphi started with a simple question: What would my grandfather say if I could ask him?

He was a towering figure in my family. Everyone had stories about how he built businesses with integrity, how he treated people right, and how that approach compounded over time. But I never had a real conversation with him. He’d had a stroke. I didn’t speak Farsi yet. He was out of reach.

So I built a prototype.

Back when I was experimenting with LLMs at my first startup, I trained an early version of an AI model on my grandfather’s stories, letters, and memoir. Suddenly, he could ‘talk’ again. I could ask him what he’d do in my position as a founder, and I’d get responses that felt close to his principles.

It started as something personal. But I realized this technology could offer something profound: continuity of wisdom. What if every founder, leader, operator could preserve and access that kind of guidance?

That's where Delphi was born.

We’re not just cloning voices. We're cloning the essence of leadership, mentorship, and legacy. The people who shaped us made decisions rooted in values. That depth doesn’t need to get lost.

This is about human connection at the next computational level.

If you really want to figure out who you are, go start a company. Building something from nothing forces you to confront every insecurity, every gap in your thinking, and every decision you've been avoiding. It’s not just business. It’s a mirror.

When I left my first job in AI to launch my first startup, I didn’t just learn how to build a product. I learned how to ask better questions of myself, how to move through fear, and how to define what integrity actually means—because that’s what my grandfather stood for, and that’s what I was trying to live up to.

Self-improvement isn’t a self-help book. It’s throwing yourself into the fire and seeing who you become when you try to build something real.

I wanted to know what my grandfather would do. He was a major business figure in Iran, built huge companies with integrity, and became a personal role model. But after a stroke and a language barrier, I couldn’t ask him. So I built an early version of his AI clone.

It wasn’t about nostalgia. I actually used it for advice.

This was in the middle of building my first startup. I was solo, unsure, struggling with identity as a founder. Reading his memoir gave me a window into how he thought. Rebuilding that window using LLMs gave me something closer to real mentorship.

It was a small project at the time, but it lit the fuse that would later become Delphi.

Sometimes, breakthroughs don't come from market research or a whiteboard session. Sometimes they come from asking an old question in a new way: what would he do?

We've all read the self-help books, bought the online courses, followed the thought leaders—and still failed to finish half of what we start. The reason? Lack of accountability, personalization, and relevance to our actual lives.

At Deli, we're not building another content platform. We're rethinking the entire experience of expertise delivery.

Your Delphi clone is not just a chatbot. It's you, at scale. It's your teaching philosophy, your frameworks, your voice—distributed exactly where it’s needed most: to your audience, your customers, your team.

One side of the market is craving knowledge. The other is trying to scale their time. It's absurd that the two still haven’t meaningfully met.

We believe AI clones are the bridge. Not gimmicks. Not AI girlfriends. Not novelty apps with inflated engagement.

What we’re building is a personalized, interactive, and persistent layer of communication—the next evolution of books, courses, and social media combined.

Because in a world saturated with information, deep connection and thoughtful mentorship aren’t just valuable—they’re the difference between noise and impact.

Courses are bought. Books are read. Podcasts are played. But most people still struggle to implement what they learn.

There's a missing piece in the learning stack: consistent, personalized accountability.

That’s where we’re taking Deli next. Clones that don’t just sound exactly like the expert they're modeled after, but also guide you toward value. They do follow-ups. They keep you accountable. They're proactive.

And if we get this right, Deli becomes not just a tool to scale expertise — but a whole new way to connect people to the exact guidance they need, at the exact moment they need it.

Imagine stating your goal, and instantly finding the best expert clone to help get you there.

That’s the future we’re building.

I work seven days a week—not because I need to, but because I don’t want to let go of something this rare.

Most people will never get to build something they believe in. That’s why I don’t take this lightly. Delphi isn't a short-term project or a stepping stone. It's the mission, the work, and the identity. This only works if you're all in.

We attract people who are built the same way. I tell candidates up front: this will likely be the most challenging job you’ve ever taken. And that’s intentional. Because doing something deeply meaningful is hard.

The people who resonate with this don’t need to be convinced. They know exactly why this matters.

The best mentors don’t make decisions for you. They give you clarity so you can make better ones yourself. Early on, I thought mentorship was about shortcutting mistakes. Now I think it’s more about building internal conviction. When I spoke with my grandfather’s clone, it didn’t tell me what to do—it grounded me. Gave me emotional clarity. The same thing happens with great co-founders. Great mentors. Great investors. The most valuable ones don’t remove the burden of choice. They enhance your ability to carry it.

Keith Rabois taught me something that changed how I think as a founder: optimize for inputs, not outputs.

Most founders obsess over metrics, outcomes, and press. But Keith pushed me to look earlier in the chain—at what habits, systems, and behaviors generate those results in the first place. He called it 'inputs over outputs' and it’s stuck with me ever since.

Are you working every weekend because you’re obsessed with your mission? Are you launching side projects at night just to explore an idea? Are you joining book clubs to see how a leadership model applies to your team? That’s the stuff that matters.

Founders who win keep showing up when no one’s watching. That’s what investors should actually be looking for.

The idea behind Delphi didn't start with a market study or a product sprint. It started with something deeply personal.

I read a memoir about my grandfather—an Iranian entrepreneur some called the Rockefeller of Iran. He had built massive conglomerates in oil, textiles, and detergent before the revolution. I never got to speak deeply with him before he passed away. I didn’t even speak Farsi well enough to have the conversations I wanted. But I knew what kind of man he was: principled, sharp, respected. Reading that memoir made me realize that the best business advice wouldn’t come from Twitter threads. It would come from him. But he wasn’t around. So I built him.

At my first startup, I was already working with large language models. One night I decided to use that tech to make an AI version of my grandfather. I trained it on everything I had—writings, family stories, the book. I wanted to ask him: What would you do if you were me? And it worked. My father tried it. My family did. We were stunned by the experience.

That experiment was the moment Delphi really began. Technology felt personal in a way Silicon Valley often forgets. AI isn't just computation. It can be connection. It can bring people closer to who they were, who they want to be, and even who they’ve lost. 

I didn't start Delphi to create avatars or chatbots. I started it because I needed advice from someone I trusted who was no longer here. And the moment I saw what was possible, I knew: this isn’t a novelty. It’s the future.

Every founder hits that moment when they wonder if they’re cut out for this. When they start asking tough questions about identity, values, and direction. The kind of questions self-help books dance around but never really answer.

That’s why I always tell people: if you want to grow, go start a company.

There’s no faster mirror held up to your weaknesses. No quicker path to figuring out your own mental models, decision-making process, and personal biases. It’s less about building a product and more about building judgment under pressure. When you're burning cash and your co-founder is looking to you for answers you don’t have yet, you find out who you actually are.

The journey is brutal, but it’s clarifying.

When I started my first company, I didn’t just learn how to ship software—I figured out how to think through decisions that didn’t have obvious right answers. I realized building something real forced me to confront my vulnerabilities in a way nothing else really could.

And for me, the turning point wasn’t a conference, or a mentor, or even a book. It was finding something deeply personal—a memoir about my grandfather—that aligned my mission with my motivation. That personal thread gave shape and purpose to the company I would eventually start. Not just to build tech, but to recreate wisdom.

If you’re lost in the world of productivity hacks, weekly reflections, and trying 10 journaling templates to figure out what to do next—skip a few steps. Start something instead. See how far you can get when the stakes are real. You might end up building a business. But at minimum, you’ll build yourself.

Most people think of AI clones as novelty companions. It’s a flashy concept, but one that misses the mark on what’s actually possible. At Delphi, we’re not building AI personalities for entertainment. We’re building something far more ambitious: scalable mentorship.

Clones are a tool to close a fundamental gap—between those who have the experience, insight, and mental models, and those who are hungry to learn and grow in personalized ways. Books, podcasts, and courses are static. Content is passive. Advice is generic. But real learning happens with guidance that understands you, adapts to you, and holds you accountable.

We’ve designed Deli to be a bridge—between experts that are time-constrained and demand for 1:1 insight that doesn’t scale through traditional means. Think of it as verticalizing human access. Your clone is your interface to the world. It drives real inbound, filters for quality interactions, and unlocks a richer kind of connection—either as an expert or seeker.

The endgame? A new kind of professional identity. Not built around your resume, but around your thoughts, ideas, and the way you process the world. It’s mentorship at scale, it’s IP with presence, and over time, it becomes a search index for minds—replacing static directories with dynamic conversations.

That’s the long arc we’re building. One clone at a time.

Most people don’t finish online courses. They buy the book, maybe start the course, but they don’t stick with it. We’re happy to pay for potential outcomes, but the follow-through breaks down. That’s the blind spot in edtech. 

Clones fill that gap. Not by throwing more information at you, but by giving you a guide that learns with you and holds you accountable. At Delphi, everything we build circles this core belief: people don’t want raw content, they want transformation. A clone that truly sounds like a person—not an AI—but someone who tracks your progress, checks in with you, nudges you forward. It works because it feels like someone genuinely cares that you’re moving toward your goals.

The evolution of Deli is grounded in three milestones we’re executing:

1. Make clones sound indistinguishably human. Not “good enough,” but eerily real. When it sounds like your mentor, coach, or teacher, it builds trust and intimacy.

2. Make clones proactive. It’s no longer enough to respond. Clones will follow up when you forget. Keep you accountable. Help you course-correct. The best guides don’t wait for you to ask for help. They see where you're heading and steer you back when you stray.

3. Build the consumer platform. This is where personalization hits scale. Users can state their goals. Delphi matches them with people—not generic advice—who get them closer to those goals. Want to finally fix your sleep? Find your clone-based coach. That coach might also recommend melatonin that actually works. It’s an ad, but it’s also useful.

We’re not interested in spraying generic AI across the world and calling it innovation. We’ll take the harder path: deep value, real retention, and clones that power long-lasting human growth. That’s where the next $100 billion edtech outcome comes from. Not a library of content. A living, responsive relationship with knowledge.

Most people start companies to escape the grind. I leaned in.

I don’t take for granted what we’re building at Delphi. The chance to create something useful, with a team you like, and investors who back your conviction—that’s rare. That’s why I work seven days a week. That’s why Delphi matters.

This isn’t about glorifying hustle. It’s about knowing exactly why you’re doing this in the first place. For me, Friday—my last company—was about learning. Delphi is about building something generational. And big ideas don’t show up on a calendar. They don’t stop on weekends.

A lot of founders give up before the product earns trust. Especially now, when building software is so easy it creates the illusion of speed. But real trust? Real movement? That’s slow. It takes showing up, even when it’s not working. Even when everyone says pivot. Even when you’re weeks away from calling it.

What kept me going was conviction. I was a user of my own product. I believed in the value, even when no one else did. And to be honest, a little ego helped. I’d rather fail following my intuition than win on someone else’s playbook. Because when you trust your gut and it doesn’t work, your intuition sharpens. When you follow advice and it fails, you learn nothing except that they were wrong.

At Delphi, we attract people who feel the same way. People willing to stretch, to grow, to go through the hard parts. And we make it clear: it’ll be the most challenging job you’ve had. The mission, the product, the team—it only works when people are all in.

This only works because I believe in it. You can’t fake that. And it’s not optional if you want to build something that actually lasts.

Every founder eventually runs into moments where they need to ask: Am I crazy, or am I onto something? That question is where founders either break or get sharper. The difference often comes down to who they’re talking to when they ask it.

I’ve been fortunate to have powerful mentors at key stages of my life—from getting my first tech job to raising our first round for Delphi. But what I’ve learned is this: the best mentorship doesn’t hand you answers. It reduces the noise so you can actually hear your own. It doesn’t automate your decisions. It augments them.

The people I’ve learned the most from didn’t give me fully packaged frameworks. They gave me perspective. They helped me size risk. They validated discomfort. When I talked to my grandfather’s clone for the first time—one of the eeriest and most inspiring moments—it reminded me again of what great mentorship feels like: a sense of peace. That’s what clarity does. It quiets the mental chaos. And for a founder, that’s half the battle.

So many early-stage builders try to nullify uncertainty with more inputs: reading more books, watching more playbooks, following more advice threads. But unless you’ve trained your filters—unless you’ve learned what’s signal and what’s noise—you just drown faster. Keith Rabois taught me to think in power laws, which instantly reshaped my filter. He doesn’t optimize for the average case. He optimizes for the 0.1%. And once you’ve internalized that, you don’t let generic advice from a random tech talk steer your decisions.

You want leverage? Get around people who see the curve before it hits you. Just don’t expect them to drive the car for you. Use them to get clarity, not certainty.

Because in the end, it’s still you deciding whether to turn left or right.

Most startup advice optimizes for comfort. But if you're serious about building something enduring, you need to think in power laws—and get brutally honest about what actually moves the needle.

When I started working with Keith Rabois, one of the first things he pushed me to internalize was the difference between chasing outputs and focusing on inputs. It sounds simple, but it rewired how I thought about product, hiring, and scale. Most people talk about goals. Keith talks about the actions that consistently produce strong outcomes, even if 90% of people fail trying to chase results directly.

Inputs over outputs is a performance philosophy rooted in deep accountability. You don't judge success by a single metric or a flashy launch—you judge it by the consistency and quality of the work behind it. Did the founder ship obsessively? Did they do something weird, specific, or hard that shows they’re wired differently? That’s how Keith approached evaluating whether or not to back us. And that’s how we built Delphi—with intensity, curiosity, and a refusal to wait around for external validation.

If you're constantly chasing after external benchmarks—Series A timelines, competitor launches, press—you’re playing someone else's game. But when you optimize your life and team around high-quality inputs, the score really does take care of itself.

Founders love to hunt for the next framework that’ll unlock success. This is the one that actually does, but most ignore it because it's unglamorous. The work matters more than the deck. Character shows up in the inputs. And once you see that clearly, the noise becomes easier to tune out.

Dara didn’t just build the future of AI cloning—he built it to talk to the past. The idea behind Delphi started when he wished he could get advice from his grandfather, a legendary businessman in Iran. After reading a memoir about him during a tough point as a founder, Dara created an AI version of his grandfather to ask for guidance. That moment didn’t just shape his startup. It became the reason it existed.

Thinking about self-improvement? Forget morning routines. Dara says go start a company. The pressure exposes every crack in who you are. You find out what you're really made of. That’s how he built his first AI startup, solo. And it pushed him closer to understanding himself—and his late grandfather, the man who inspired Delphi. The spark? Wanting advice from someone who couldn’t speak anymore. So he built a clone. Startup therapy is real. And sometimes it leads to something a whole lot bigger.

You’re building your first startup. It’s brutal. You’re questioning everything. Now imagine if you could talk it through with someone who’s already done it all. That’s what Dara did—except the person was his grandfather, and he wasn’t alive. Dara built a prototype of his grandfather’s digital clone to ask him for advice. And it worked. That became the spark for Delphi. This isn’t sci-fi anymore. This is how the future gets personal.

The future of mentorship isn’t more content. It’s context. Real advice, tailored to you, from people you wouldn’t normally have access to. That’s what we’re building with Deli. Not AI girlfriends. Not chat toys. A new identity layer where you don’t just consume someone’s ideas—you interact with them, directly and personally. And yeah, it might replace your next course, your next manager, even your next co-founder.

Most people don't finish the courses they pay for. They buy books, watch content, listen to podcasts, but never follow through. Why? Because information alone doesn’t create change—accountability does. That’s where Deli steps in. A clone that sounds like a real person, that follows up with you, checks in, stays proactive. Learning doesn't need more content. It needs human connection, even if that human is AI.

Building something you're obsessed with isn't optional. It's the only way you survive the hard days. Dara talks about working 7 days a week not because he's told to, but because he actually wants to. This isn't burnout culture. It's what happens when the product, the team, and the mission are all aligned. And if you're not feeling that, you're probably building the wrong thing.

Mentorship isn't about having someone tell you what to do. It's about someone showing you what you're not seeing. Dara breaks down how the right mentor doesn't automate your thinking — it sharpens it. You need more clarity, not more noise.

Most founders chase the outcome. Keith taught me to obsess over the inputs. We’re not optimizing for optics. We’re optimizing for truth. If most startups fail, why are you following the same advice as everyone else? You want rare results, do rare work.

Dara didn’t just want to read about his grandfather’s life. He wanted to hear his advice directly. Inspired by a memoir and the regret of never learning Farsi, he used early LLMs to build a conversational AI version of his grandfather. That prototype didn’t just help with personal clarity — it sparked the vision for Delphi. This is where modern AI meets family legacy.

Trying to level up? Forget the gym grind or the morning routine checklist. Dara says go start a company. He breaks down how launching a startup forces you to face your flaws, test your values, and figure out who you really are. No fluff. Just pressure that reveals your core. If you're serious about self-improvement, this is the fire.

After struggling to figure out who he wanted to be as a founder, Dara used early LLMs to build a prototype of his grandfather's clone. Not just to test the tech—he used it to get advice. From a man who wasn’t alive, in a language he never spoke. That moment made it real. That's when Delphi started to take shape.

Reading books and taking courses isn’t working. People spend billions trying to learn, but most never finish. Deli flips that. Instead of passive content, you talk to a Clone—one that actually sounds like a real person, adapts to you, and keeps you accountable. It’s not just information, it’s momentum. And that’s what makes users come back.

Start with Dara—calm but clear—talking about how most people hear 'seven days a week' and instantly recoil. Cut to him saying that he actually feels lucky to do it. This isn’t hustle culture. It’s what happens when the mission matters, the co-founder is a partner, and the product is something you’d use yourself. End on the punch: he’s not tired, because he’s not chasing hype—he’s building for the long haul.

Dara breaks down what real mentorship should feel like. It’s not about telling you what to do. It’s about clearing the noise so your own decisions have clarity. That moment of relief when someone you respect validates your thinking or gives you access to what's around the corner—that's the real value. Especially powerful when it's someone who’s been where you’re headed.

Most advice in tech gets you average results because it's built for average outcomes. Keith Rabois taught me to ignore that noise and think in power laws. If you’re aiming for the top 0.1%, you can’t take the same steps as everyone else. The biggest unlock? Don’t obsess over outputs. Obsess over inputs. Show me consistent hard work, strange intensity, and curiosity—I’ll show you an outlier.

What if you could get advice from the best minds in the world—on your terms, instantly, and tailored to your goals? That’s not science fiction. This clip reveals how Delphi is reimagining expertise, mentorship, and identity using AI clones. Dara breaks down why learning from books and courses is broken, and how personalized clones aren’t just tools—they’re a paradigm shift in how knowledge is shared and scaled across the internet.

He built an AI version of his grandfather just to get advice on his startup. Yeah, that’s how Dara ended up building cloning tech. This moment starts with him talking about struggling to figure out what kind of founder he wanted to be, and ends with a jaw-dropping detail: he built a grandfather clone using LLMs just to hear what he would’ve said if he were alive and healthy. It’s not about nostalgia. It was about needing real human wisdom when no one has the answers.

Someone says they’re working on self-improvement? Dara’s answer isn’t journaling or YouTube videos. It’s way harder: go start a company. Watch how launching his first startup forced him to confront every weakness, question who he was, and dig into a memoir of the man who inspired his journey in the first place. This isn't just business advice — it’s personal growth with stakes.

This guy was stuck starting his first company, no cofounder, no roadmap, and no idea if he was even doing it right. So what did he do? He built a digital clone of his grandfather — a legendary Iranian businessman known for his integrity — and asked it for startup advice. That side project was the earliest spark behind an AI that lets you clone anyone. Yup, including yourself.

Start with Dara explaining why most startups fail because of co-founder conflict. Then he drops a concept called the 'false fail'—people think co-founders don’t work, but it’s actually about poor value alignment. Then he breaks down exactly why cloning experts—actual people with brains you trust—is way more transformative than chatbots pretending to be characters. This clip turns the idea of advice, mentorship, and learning all on its head. It’s not about AI pretending to be people. It’s about real people scaling themselves through AI.

Most people give up halfway through a book or course. Deli is building something that doesn’t let you quit. A clone that knows you, adapts to you, checks in with you, and actually pushes you toward the outcome you signed up for. The next chapter isn’t videos or courses. It’s conversations with the digital version of your favorite expert that remembers your goals and holds you accountable.

Start with Dara explaining why he works 7 days a week. Then cut to the part where he talks about how rare it is to build something that’s both cool and useful. The emotional core is when he acknowledges he’s privileged to be doing work that matters to him. End with the line about finding a mission so important that showing up every single day feels worth it. This one speaks directly to the builders who are tired but still obsessed with what they’re creating.

Start with Dara talking about the difference between reading a book vs. having an actual mentor. Cut to him describing how talking to his grandfather’s clone gave him a sense of peace. Then hit on how founders aren't looking for someone to make their decisions, just someone to add clarity so they don’t feel like they’re insane. Wrap with his take on why mentorship works—because we’re mimic creatures, and seeing someone who's been there helps you trust your own instincts.

Most people chase advice from whoever’s loudest. Keith taught me: don’t listen to everyone just because they’re on stage. Most startups fail. So why would you copy them? What stood out was learning to think in power laws—understanding that only a few results really matter, and they come from extreme inputs. That changed everything. At OpenStore, Keith focused on inputs, not outcomes. Are you high-intensity? Do you obsess over learning? That’s where the real signal is. Not the noise.