Keep Going

When you talk to Diane Yu, you hear the mix of humility and steel that makes a founder stand out. She’s the co-founder and CEO of Tidal Wave, an AI company that is trying to modernize one of the most entrenched and slow-moving sectors out there: mortgages.

Her take is simple. We live in a world where everything else moves fast—food, rides, even love—but mortgages drag on, slow and archaic. Tidal Wave is her answer: an AI-powered mortgage point of sale system that does the hard work of evaluating and completing applications before an underwriter even looks at them. It’s not flashy, but it’s practical. And in this industry, practical is revolutionary.

Most of us know Anker as the little black brick that saved our phones when the battery dipped into the red. What’s less known is how Steven Yang built the company. He started as a Google software engineer, left a stable life, and moved into hardware without any real experience. That leap, and the mistakes along the way, tell us a lot about what it takes to keep going.

1. Curiosity is stronger than fear

Steven admitted that if he had known how hard it would be to move from software to hardware, he might not have done it. But he didn’t overthink it. He saw that early smartphones had weak batteries, looked at the clunky packs people were carrying, and thought, “I can do this better.” Sometimes progress is just that simple: ignoring the warning signs and trusting your curiosity.

2. Build in steps, not cliffs

Anker didn’t try to be Samsung overnight. They went from replacement batteries to portable chargers, then to cables, headphones, and finally smart home devices. Steven called it “climbing stairs, not cliffs.” That lesson applies anywhere: if you can break a massive risk into smaller, connected moves, you have a better chance of surviving.

3. Failures are tuition

The company’s first big stumble came with their 3D printer. Engineers decided to use a dual USB-C cable to connect the moving print head. It worked fine at first. Then months later, customers found the connection failed during long prints. Anker had to recall units. For Steven, the pain wasn’t just technical—it was knowing that loyal customers were frustrated. The fix wasn’t to hide the mistake, but to absorb it like tuition, improve testing, and write down new rules so the same mistake wouldn’t happen again.

4. Keep the customer at the center

When a product breaks, the instinct is often to defend yourself. Steven said the instinct at Anker is different: first, take care of the customer. Then fix the process so it doesn’t repeat. That order matters. It’s what keeps people trusting you, even when you slip.

5. Think long term, but keep tinkering

Steven talked about edge AI—models that live on devices in the home, not in the cloud. His vision is layered: massive models in the cloud, medium models in the home, tiny models in wearables. They’ll all talk to each other. It sounds futuristic, but he insists it’s just a few years away. What stood out wasn’t the prediction, but that he’s still tinkering like an engineer, still trying to solve problems from the ground up.

6. Build platforms, not empires

When I asked if he would have been happier staying at Google, he said no. What drives him now isn’t just making products, but creating a platform where other makers can build too. That’s the deeper lesson: success isn’t about being the smartest person in the room. It’s about giving people the tools and space to create something better than you could on your own.

Steven Yang left a secure job, stumbled through hardware, recalled products, and still came out with one of the most trusted consumer brands of the past decade. His story isn’t about charging bricks. It’s about curiosity, humility, and the discipline to learn from mistakes without letting them stop you.

I sat down with Lisa Gralnek, Managing Director for the U.S. and Global Head of Sustainability and Impact at iF Design. She also hosts Future of XYZ podcast. We talked about awards, values, and the hard work of keeping design honest.

Here is what matters.

iF Design is not a new project. It started in 1953. It is a nonprofit in Germany. The award is large, global, and serious. Entries come from scores of countries each year. Jurors come from real jobs, heads of UX, design chiefs, leaders who ship. The process runs in two stages, an online cut, then an in-person review in Germany. Every entrant gets a score map, even those who do not advance. Idea, form, function, differentiation, and sustainability. That last one is now a fifth of the grade. This is rare. It makes the feedback useful, not fluff.

Why enter at all? Because clear, external critique is hard to get. Because you learn where you stand in your field. Because a public win travels, and a public loss still teaches. Most award mills take the fee and send a badge. This one sends a readout you can act on.

iF is pushing into the U.S. now. Lisa was hired to build that bridge. The brand is known in Europe and Asia. Here, less so. Growth has been slow and careful. Small budgets. More community than ads. More teaching than hype. They launched the iF Design Academy to share what the field has learned, and to help designers learn what schools often skip. Personal leadership. real future literacy around sustainability and new tech. most of all, business fluency.

That last point hit home. Designers have too often ceded the boardroom by speaking only in color and taste. If you want a seat, learn gross margin, capex, payback, risk. Learn to defend a choice with numbers. Learn to change a choice when the numbers say you should. Beauty without a model does not survive a budget review. Beauty with a model can.

Values are the spine of the institution. Lisa was blunt. iF holds to impact and excellence. It funds student and social prizes. It set up rules to keep quality high when volume rose. It added an early round to filter weak entries. It made sustainability part of the score. This is how you stay on course in a long life. You write the rules down. Then you keep them when it costs.

We talked about AI. I asked the old question, what happens when a prompt can spit out a chair in a famous style in minutes. Lisa did not hand wave. The path is clear. Use AI where it is a tool, drafts, planning, research. Keep the human where it counts, framing, taste, judgment. Keep your scope tight. Be open about how you used the tool. Add safeguards. Do not let your work flatten into sameness. AI can speed the task. It cannot carry the meaning.

Brand work came up as well. How do you grow something that feels faint in a loud market. You tell the truth about what you do. You back winners with real exposure. You bring jurors and teachers into the tent. You keep showing up. It takes time. That is fine. Things that last usually do.

If you are a founder or a head of design, here is my ask. Enter where the feedback will help you ship better, not where a sticker will pad a deck. Write your team’s values in plain English. Share a code of conduct with clients. Teach your designers the language of money and risk. Teach your leaders the language of people and care. Use AI with a plan, and say what that plan is. Build a product that can face a jury and a P&L, and still feel human.

One more note. The iF site holds a record of winners back to the mid-fifties. Spend an hour with it. You will see a clear line from then to now. Form, function, care. Dieter Rams stood on their stage and reminded a full hall that designers have a duty to make the world better. That is not a slogan. It is a job description.

On the latest episode of Keep Going, I spoke with Neil Markey, CEO of Beckley Retreats, about how he went from combat deployments and corporate burnout to guiding people through psilocybin retreats in Jamaica and the Netherlands.

Carrying Forward Amanda Feilding’s Legacy

Beckley Retreats is the living continuation of Amanda Feilding’s work. For decades, Feilding ran the Beckley Foundation, a nonprofit dedicated to psychedelic research and drug policy reform. She’s been called the hidden hand of the psychedelic renaissance — lobbying governments, funding research, and pushing for policy change long before the mainstream was ready.

When she passed away in May at the age of 82, her family and collaborators, including Markey, stepped forward to carry her vision into practice. Beckley Retreats is that next step: creating safe, structured programs to bring psychedelic experiences to more people.

From War to Burnout to a Different Path

Neil’s path to Beckley is not what you’d expect. He joined the military after September 11, serving in Iraq and Afghanistan. After leaving the service, he pursued graduate degrees in business and international affairs, then went into the corporate world. On paper, it looked like success. In reality, it felt like collapse.

Sleepless nights, short temper, alcohol creeping into the routine, relationships fraying — the same symptoms that haunted him after combat were back in boardrooms. Twice in his life, Neil found himself on the edge, burned out and discontent.

What changed? He stepped away. He pared down expenses, saved enough to create breathing room, and allowed himself to sit in stillness. That space made it possible to find new direction.

Building Retreats with Structure and Care

At Beckley, the retreats are designed as full programs rather than isolated experiences. Guests go through digital preparation first: meditation, breathwork, yoga, nutrition, and group connection. Then comes the five-night retreat in Jamaica or the Netherlands, where participants take part in two guided psilocybin sessions, held in nature with live music and careful facilitation.

Afterward, integration lasts six weeks. This is where real change can take root. With the brain and body in a neuroplastic state, new habits — daily meditation, mindful eating, deeper self-care — have a chance to stick.

The results Neil has witnessed are striking. Guests report more gratitude in daily life, more patience in relationships, clarity about decisions, and sometimes even a quiet end to habits like drinking. One story stuck with him: a fellow veteran whose young son had always chosen his mother for comfort suddenly began coming to him at night after the retreat. Something shifted, not just in him, but in the way his family felt him.

Choosing Work That Matters

Neil laughs when he admits he’s back in an intense role — running retreats is far from easy. But this time, the effort feeds him. The work aligns with his values. As he told me, it isn’t about escaping hard work, it’s about finding work worth doing.

That’s the heart of Beckley Retreats: creating space, breaking cycles of burnout, and offering people the tools to reset their lives. And it’s fascinating to meet the guy who is helming it all.

When I started Keep Going, I wanted to talk to people building things with care. Not the VC-fueled wrecking balls chasing unicorn status, but the ones trying to solve real problems without leaving a crater behind. Nicki Sprinz, CEO of digital product studio ustwo, fits that bill.

In a startup world that still fetishizes disruption, Nicki is pushing a different message: stop disrupting, start caring. And it’s not just a tagline. It’s the operating philosophy of an international company working at the intersection of technology, wellness, and responsible design.

Here’s what I learned from our conversation.

Keep Going - A Guide to Unlocking Success is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Disruption had its moment. This is something else.

Remember “move fast and break things”? Nicki does. But when you’re building tools for health, behavior change, and everyday life, that mentality doesn’t cut it. “You don’t get to break things when patients are involved,” she told me. You slow down. You design with empathy. You build with rigor. Not everything should be frictionless. Sometimes the point is to add friction—the good kind that helps people pause, reflect, or make a better choice.

It’s not just about making apps that work. It’s about making apps that should exist.

You can have engagement without addiction.

We got into the difference between engagement and addiction. The former keeps people coming back because it’s meaningful. The latter keeps them trapped. At ustwo, they’re focused on emotional design. Think Monument Valley—the studio’s breakout game that felt like a meditative puzzle box—not the dopamine death loop of infinite scroll.

Design, Nicki says, can invite autonomy. Good design earns your attention. It doesn’t hijack it.

AI isn’t replacing us. It’s sitting next to us.

Nicki had a nuanced take on AI. Yes, it makes things faster. Yes, it can help with structure, drafting, and research. But it can also make everything look the same if you’re not careful. At ustwo, they’re experimenting with AI tools, but always with human context and judgment at the core. They’re designing “adaptive AI”—systems that respond to users in thoughtful, helpful ways. Not just content machines, but collaborative tools.

They even built Sproutiful, a project that used multimodal AI to help people track their nutrition and make small behavior changes. The idea was to nudge, not nag. And it worked.

Culture isn't a slide deck. It's who you fire.

One of the most powerful parts of our talk came when I asked Nicki about firing a client. She didn’t dodge. She told me about a time they walked away from a big project because the client’s behavior clashed with their values.

At ustwo, those values aren’t decoration. One of them—Be Human—means don’t shout, don’t bully, and treat people with respect. It’s written into their code of conduct. It guides their hiring, their projects, and their client relationships. And when someone violates that, they walk.

That kind of consistency is rare. It’s also what holds a company together when the money gets tight.

The AI future still needs apprentices

We talked about the hard stuff too. What happens when AI eats the entry-level jobs? How do you train the next generation when there’s no “junior” left? Nicki doesn’t have all the answers, but she’s thinking hard about it. “We still need people in the room together,” she said. “Even if AI is joining us in that room.”

The human element—mentorship, nuance, values—isn’t going away. At least not if companies like ustwo have anything to say about it.

So yeah, ustwo is a product studio. But it’s also something else: a test case for how to build tech in a way that doesn’t feel gross. That doesn’t chew people up. That doesn’t mistake velocity for vision.

Nicki didn’t flinch when I pushed. She gave honest answers about tradeoffs, mistakes, and hard decisions. She talked like a human being, not a CEO out of central casting. And that, more than anything, made me think she’s onto something.

We live in a world obsessed with velocity. Founders are supposed to scale fast, pivot often, and raise money like their lives depend on it. But not every success story fits that mold. Some companies are built slowly, deliberately, and on their own terms. Reedsy is one of them.

I first wrote about Reedsy back in 2014. It was a small, strange idea at the time: a marketplace for authors to find editors, designers, and other publishing professionals. Now, more than a decade later, Reedsy helps produce over 50,000 books every year. It’s profitable. It’s stable. And according to founder Emmanuel Nataf, it’s calm.

We had him on Keep Going to talk about how he pulled that off.

No deadlines. No burnouts. No nonsense.

Emmanuel’s approach to building Reedsy is almost aggressively anti-startup. He never bought into the blitz-scaling mindset. Instead, he focused on building something that worked—then waited to see what it wanted to become.

The company raised a little money early on, but quickly moved to profitability. No pressure to grow at all costs. No mad rush for the next round. Just a team of about 50 people, working quietly and consistently, building tools that authors actually want to use.

That freedom has shaped the culture. There are no arbitrary deadlines at Reedsy. No imaginary pressure. If a product feature isn’t ready, it waits. If the summer is slow, they accept it. December? Forget about it. They’re not chasing unicorn status. They’re building something real, at their own pace.

Studio: A writing app that doesn’t yell at you

One of Reedsy’s biggest moves was launching Studio, their collaborative writing platform. It's where writers can draft, plan, revise, and format their work—all in one place. It started as a humble book editor, but over time, it became something closer to a virtual writer’s room. Studio lets you build characters, map worlds, outline stories, and track your progress. You can even invite editors and beta readers to work with you in real time.

It's a calm product, made by a calm company.

Writers will invest in themselves

One of the best insights from the interview was how Emmanuel thinks about pricing and value. Writers don’t always have a lot of money, but they do invest in their work. Just like photographers buy cameras or musicians buy guitars, writers will pay for tools that help them finish the thing that’s clawing at them from the inside.

That understanding—that writing is a form of personal investment, not just a commercial pursuit—shapes the whole business.

Growth, the old-fashioned way

Reedsy isn’t stagnant. It’s growing steadily. But Emmanuel’s version of growth doesn’t rely on sprints, pressure, or panic. It’s about listening to the users, iterating with care, and avoiding hype. Even when things slow down, he sees it as part of the rhythm. "You keep digging, and usually, you find something,” he told me.

That’s how they’ve unlocked new waves of growth again and again—by being patient enough to let them happen.

A life beyond the company

In the end, what stood out most was how Emmanuel talked about life outside of Reedsy. He and his co-founders aren’t martyrs to the startup cause. They’ve built lives with room for more than work. They write. They think. They breathe. Reedsy is central, but not all-consuming.

It’s the opposite of the founder-as-hero myth. It’s just a group of people building something good, carefully.

And that, to me, is what Keep Going is all about.

When Max Buchan launched Valarian in 2020, he wasn’t building the next messaging app or enterprise dashboard. He was building infrastructure—systems that sit under the surface of apps, institutions, and, more recently, battlefields.

Valarian didn’t start as a defense company. It started as an enterprise software venture aimed at helping organizations control where and how their data moved. But the world changed, fast. Brexit exposed new tensions around jurisdiction. Ukraine reminded us that sovereignty isn’t a metaphor. And suddenly, the demand for what Valarian was offering—actual, operational digital control—moved from the boardroom to the front line.

On this week’s Startup Show, Max joined me to talk about how Valarian went from fintech roots to working with NATO governments, deploying secure infrastructure in the field, and raising money from top defense investors like Scout Ventures and Artis.

From Crypto to Compartmentalization

Max cut his teeth at CoinShares, a crypto asset manager that scaled from a tiny team to a publicly traded company. There, he saw firsthand how fragmented and vulnerable digital infrastructure can be—especially across borders. That experience, combined with the geopolitical shifts of the last few years, shaped his thesis: the next era of sovereignty would be digital.

Valarian’s core product is called Acra. It's a portable, containerized backend that enterprises and governments can deploy locally, on demand. In enterprise use cases, that means secure internal communication and controlled data flow. In defense contexts, it means spinning up a network in minutes—sometimes from a Pelican case in hostile terrain.

You don’t just own the encryption keys. You own the system.

The Infrastructure Under the Infrastructure

Acra looks less like an app and more like a replicated backend system. It runs databases, object stores, and secure communication protocols—everything you’d need to build on top of, without ever touching public infrastructure. It’s generative on deployment, meaning each instance is unique. Not even Valarian has access to customer deployments.

That idea—zero external visibility—matters. Especially when you're talking about military communications or sensitive research environments. “Visibility is the threat,” Max told me. “Just revealing who’s talking to whom can compromise an operation.”

Dual-Use, Real Stakes

The company now operates in two tracks: Valarian Enterprise, which serves clients like banks and pharma companies, and Valarian Defense, which handles secure deployments for NATO member states.

Their customers are using Acra to do everything from run compartmentalized AI workloads to spin up isolated networks in the field. Some want the system to be thermite-compatible—able to self-destruct in a crisis. Others just want to ensure their proprietary data stays in one jurisdiction and doesn’t get copied into someone else’s cloud.

That flexibility—running one system in a bank, another in a war zone—is what makes Valarian stand out in a space where most infrastructure is either too lightweight or too rigid.

What Europe Needs Now

Max is blunt about the state of defense tech in Europe. While budgets are growing, capability isn’t always keeping pace. The U.S., he says, has built a far more responsive innovation ecosystem. Companies like Palantir and Anduril have shown that dual-use startups can thrive when the government is ready to work with them.

Europe’s challenge isn’t just funding. It’s culture.

“We’re more risk-averse,” Max said. “But that’s changing. People here want to buy British software. The problem is, there hasn’t been much to buy—until now.”

Valarian wants to be part of that shift: not just building defense tech, but helping shape a sovereign, scalable digital infrastructure that Europe can call its own.

What’s Next

Valarian is hiring. They’re scaling fast to meet demand across enterprise and defense. They’re also looking at acquisitions—adding capabilities that fit into their secure-by-default architecture.

From fintech to the field, Valarian’s path shows what happens when you treat sovereignty not as a slogan, but as a product spec.

Learn more at valarian.com.

We talk a lot on Keep Going about mental health. About reinvention. About trying again when everything feels broken. But this week, something deeper came through.

Justice Leak—an actor, a seeker, and now the founder of Therapod—joined the podcast to talk about his mother, her Alzheimer's, and the way psychedelics changed their lives. Not in theory. Not in some lab. In a small town in Georgia, on a hot afternoon, with a handful of mushrooms and a prayer.

Justice wasn’t a psychonaut. He wasn’t raised in a culture of altered states. His mother sure as hell wasn’t. She was a third-grade teacher, a Christian, a Southern woman who never drank, never did drugs, and certainly never imagined herself in a “non-ordinary state of consciousness.”

But life has a way of closing doors until the only way forward is through one you never expected.

When his mom started to lose her sense of self—first her words, then her visual awareness, then her memory—Justice saw something terrifying: her waking life began to look like his own worst trip. The kind where you can’t tell what’s real, where rooms distort and shapes dissolve, where your own mind becomes the enemy.

And that’s when he realized: maybe these states weren’t just for introspection or trauma work. Maybe, just maybe, psychedelics could help rebuild the brain. Not just to feel better, but to function better.

So he did the thing you’re not supposed to do. He gave his mom mushrooms. Two and a half grams of psilocybin, ground up and put in capsules. No trippy music. No festival. Just a prayer and a garden.

The next day, she woke up singing. The stutter was gone. She was dressed. She was moving. It was temporary, sure—but it happened. And it kept happening. Each time they did a session, something came back. Not forever. But enough to keep going.

And here’s where the story gets bigger.

Because Justice realized the biggest hurdle wasn’t the medicine. It was the setting. These experiences didn’t belong in sterile clinics or chaotic forests. His mom needed a safe, beautiful, responsive space—one that supported the brain during this window of growth. One without disorienting floor reflections, with sounds tuned to 40 Hz, with curves and colors instead of walls and edges.

So he built it. Therapod. A new kind of therapeutic environment. Bio-adaptive, sensory-rich, and designed specifically for non-ordinary states. Think immersive art meets neuroscience. Think DMT trip meets nature-inspired architecture. Think real-time emotional feedback—without a word spoken.

Justice is still developing the tech, working with wearables that track emotional states through the body, feeding back into the room in real-time. But the mission is clear: if this medicine is going to reach people like his mom, it has to be more than molecules. It has to be infrastructure.

You can’t just hand someone psilocybin and hope. You need a room. A ritual. A reason. And someone to sit with you.

This episode left me shaken. Not just because it was moving, but because it made me ask: what if we’re underbuilding for the future we claim to believe in?

Justice isn’t selling trips. He’s building rooms for people on the edge—of memory, of function, of life itself. He’s doing what no clinical trial can. He’s showing what love looks like when you’ve run out of options.

You can learn more at experiencetherapod.com. But the real story is this: the next frontier of healing isn’t just the brain. It’s where the brain meets the world. And maybe—if we get it right—where the brain meets the end.

This week on Keep Going, I sat down with Miist—an artist whose story is as powerful as her music. She's the first native Chinese artist to break into Billboard’s Top 25 on the AC chart, and she’s worked with legends like Santana and Curtis Williams of Kool & the Gang. But her path to that stage started much earlier, and much harder.

Miist didn’t grow up thinking she’d be a singer. Until late 2023, she hadn’t even released a song. That changed when Narada Michael Walden—producer, drummer, and hitmaker—heard one of her early demos and invited her to meet. What was supposed to be a 30-minute meeting turned into a three-hour jam session and the beginning of her first album.

That debut album yielded “It’s Too Late to Love You,” a duet with Narada that's still on the radio across Europe. But her second album is where she’s aiming straight for the heart—with six songs built around the theme of human connection, each one tied to a 15-second action listeners can take to reconnect with others—and themselves.

Miist doesn’t hide the pain that shaped her. She spoke openly about surviving childhood abuse, the deep loneliness of growing up in a boarding school, and a life-threatening cancer diagnosis that resulted in losing 60% of her liver. That pain didn’t stop her. It gave her a reason to create.

What makes Miist’s work different is that she’s not just singing to entertain. She’s trying to start a movement. One of her recent tracks, “Could You Let Me Smile?”, inspired the launch of a nonprofit called the World Smile Initiative. It began with one tragic story—a young man who died alone and unnoticed—and grew into a global project translating that song into 15 languages. Each version is performed by a local artist in their native tongue. The goal: bring emotion directly to the listener’s heart, no translation needed.

And if you think that sounds too earnest for the music business, well, that’s kind of the point. “If I was with a label, they would never let me do this,” she said. As an indie artist, Miist has the freedom to do what she believes in. No hooks. No formulas. Just a voice, a message, and a belief that change starts small—like with a smile.

She’s also launched a podcast, Make Me Smile with Miist, where each episode focuses on one issue stopping us from forming deeper connections. Each episode ends with a 15-second action anyone can do, right then and there.

When I asked her if she thought suffering was necessary to make art, she paused. “I don’t think we need to,” she said, “but I’m grateful for what I went through. It taught me how to feel. And if I can feel what someone else feels, I can write their story too.”

That’s what Miist is doing—writing stories, connecting people, and reminding us that even the smallest effort can ripple outward.

And yes, she got me to smile. Even if just for 15 seconds.

On this episode of The Startup Show, I spoke with Katie Kuo, CEO and co-founder of Atomistic Insights, a biotech startup working to speed up drug discovery using physics-based machine learning. Born out of Georgia Tech and currently part of the Alchemist Accelerator, Atomistic Insights is trying to fix a known problem: new drugs take too long and cost too much. Think $2.6 billion and up to 15 years per FDA approval.

Katie’s background is in chemistry and drug discovery, and the company’s approach is refreshingly different. Instead of treating proteins like static objects, their system tracks how they actually move—atom by atom—in the body. That motion reveals hidden binding sites for drugs, which traditional drug design often misses. It’s like watching a lock move in slow motion until you see the exact shape of the keyhole.

Their method is already patented. The team has a working MVP. They’re raising their first round. And their goal is to go beyond software licensing to actually develop in-house therapeutics—taking their discoveries all the way to the clinic.

What makes Atomistic different isn’t just the tech. It’s the founder mindset. Katie isn’t chasing AI hype. She came from a lab, saw how slow and expensive drug discovery is, and wanted to do something about it. This is a startup driven by application, not abstraction.

So, what’s next? More partnerships, deeper simulations, and eventually, real drugs designed faster and cheaper using Katie’s system. Atomistic isn’t trying to replace drug discovery. They’re just trying to make it suck less.

You can learn more at atomisticinsights.com (or at least try to spell it correctly in your search bar).