Minn Kim is the founder of Lighthouse, a company that uses software and workflow automation to help frontier technology companies bring top global talent to the United States through the US immigration and work visa process. She previously worked as an early-stage AI investor and at a startup during COVID, where she encountered the immigration pain point firsthand through friends and colleagues. Lighthouse’s core premise is that professional services like immigration law are fundamentally operational and workflow problems — data collection, document management, and process coordination — that can be systematized and scaled with software rather than by simply adding more people.
From side project to company
Kim’s entry into immigration was accidental: she started informally advising friends and founders on visa pathways during COVID, charging a few hundred dollars per hour for Zoom consultations to walk people through what to expect. This was essentially college-admissions-style consulting for immigration.
She quickly realized the real bottleneck was not legal expertise but operational chaos — information scattered across email threads, Google Docs, SharePoint, and Drive, with no system for tracking or managing cases.
She began “sticky-taping” together low-code workflow automations (Airtable, Typeform, Google Docs) to systematize the process, with the explicit goal of automating herself away.
The release of ChatGPT was a pivotal moment: she realized that much of what she was doing could be done by a machine more consistently and consistently, freeing her to spend time with customers.
Two turning points made her realize this was a real business:
Within the first three months, people were willing to pay for her consulting — validating that someone would pay for what she was offering.
She eventually priced an end-to-end immigration service at six figures, and someone paid it. That was the moment she felt there was something genuinely valuable that could be scaled.
The “services as software” thesis
Kim’s core insight is that professional services don’t have to scale linearly with headcount. Most of the US economy is information services — legal, accounting, advisory — where processes live in people’s brains. Once you decouple those processes into component parts, you can make knowledge work function like software.
Traditional immigration law firms grow by adding more people: solo practitioners cap at 10–20 people, mid-size firms at a few hundred, and large firms at thousands — all following the same “add people” model.
Lighthouse applies software to break down immigration workflows (data collection, document review, document creation, case management) so that the business can scale without proportionally adding headcount.
Founder fit and the “shape of the business”
Kim distinguishes between complicated problems (known endpoint, many variables, but solvable through systematic optimization — like Uber, Tesla, Amazon) and complex problems (unknown endpoint, heavy R&D, like biotech or AGI). She sees herself as suited for complicated problems.
She describes company building as a mirror: it amplifies your strengths (she discovered she’s great at recruiting, systems thinking, and trust-building) and also forces you to confront your weaknesses (she identified that she struggles with delivering negative feedback and needed to improve).
Her advice: invest in your spikes, don’t sand them down. The things that make you unique as a founder are assets. But you should also be aware of your non-spikes — areas where improvement would meaningfully help you or the business.
To identify blind spots, she sent an anonymous 30-question Google Form survey to about 20 people (friends, former collaborators, a former executive coach) asking them to rate her on a 1–5 scale across various attributes. The answers were surprisingly consistent. One insight: she tends to see the best in people, sometimes delusionally so, which means she doesn’t always recognize when her problem-solving observations are received as criticism by others.
Lessons from co-founding vs. going solo
Kim’s first company was co-founded with a best friend. Key lessons:
Have hard conversations upfront — co-founding is like getting married, and affinity from friendship can cause you to skip necessary alignment conversations.
Friendship does not automatically translate to co-founder compatibility. People make implicit assumptions about friends’ values and working styles that they wouldn’t make with someone they just met.
When co-founding with someone less familiar, you tend to be more curious and objective about their decisions rather than assuming you already know why they did something.
She went solo for Lighthouse partly because she didn’t have a co-founder option at the time, and she realized it wasn’t a blocker. Her first two hires were engineers she already knew and trusted, which was critical for de-risking the early team.
She believes you go faster solo but farther together — a co-founder provides emotional partnership and shared investment that a solo founder has to build through other relationships.
Recruiting philosophy: long games and “try before you buy”
Kim is a strong believer in playing long games with recruiting. Her first engineering hire was a friend who wasn’t looking for a job and actually wanted to start his own company. Rather than trying to convince him otherwise, she stayed in touch over months — inviting him to co-work, to dinner — until the timing was right. He started as a contractor, went part-time, and eventually joined full-time.
She maintains a short list of 5–10 people she’d love to work with and periodically keeps them updated on what she’s building.
Contracting as a “try before you buy” mechanism has been central to her hiring approach:
It lets potential hires see “under the hood” — how the team collaborates, communicates, and operates — which is often the biggest pull for joining.
It builds trust before anyone makes a full-time commitment.
It’s also practical: early-stage companies often can’t afford full-time hires for every need.
The mission: talent infrastructure for frontier companies
Lighthouse’s mission is to build talent infrastructure — starting with frontier technology companies (AI labs, semiconductor companies, deep tech) and eventually expanding to all companies.
The core problem: the best person for a highly specialized role (e.g., a niche semiconductor engineer in Australia, of whom there may be only ~20 in the world) is often not in the US. The difference between getting that person started in weeks versus months is enormous given the pace of technological change.
Kim has met brilliant researchers at conferences like NeurIPS who had never even considered coming to the US, largely due to anxiety about the visa process and fear that checking the “needs sponsorship” box would get their application filtered out.
Lighthouse is also building data signals about where great talent comes from, what they studied, and where they work — enabling proactive talent discovery rather than reactive recruiting.
The broader vision: pulling the future forward by enabling the companies building transformative technologies to assemble the best teams in the world, faster.
The case against and for solo founding
The case against solo founding: If you’re uncomfortable asking for and receiving feedback, solo founding will be very hard. Without a co-founder who is equally invested in your growth, you have to be much more intentional about soliciting feedback from your team, customers, and advisors. You won’t grow as fast without that built-in devil’s advocate.
How Kim processes feedback: She prefers written feedback (both giving and receiving) because it allows her to process it more objectively. She doesn’t usually process feedback in real time — she reflects on it during long walks or weekend retrospectives, asking “why” to discern what to integrate and what to discard. She’s also built a retro feedback culture at Lighthouse that focuses on judgment-free analysis of what happened versus what was expected.
The case for solo founding: Kim never thought of herself as a founder before Lighthouse. When she started, she simply didn’t have a co-founder option and decided to do it anyway — and it worked. Her argument is that a co-founder is not a required ingredient to solve a problem you care about. If it was true for her, it can be true for many other smart people who haven’t considered it.
