How I Think, What I’d Back, and Why
Most capital is spray and pray- fighting over the same handful of overvalued deals because the market cannot identify which companies are capable of maintaining competitive advantage as AI accelerates the rate of innovation exponentially.
My lens is more disciplined. I look for market dominance that is structurally inevitable, commercial relationships that are stable and repeatable, technology that is near impossible to replicate or out evolve, and founders who know exactly how to compete under shifting market conditions – with grounded command and a genuine positive impact on humanity.
I have spent nearly two decades building ventures at the intersection of AI, healthcare, and frontier technology. Those years gave me the ability to ask questions that business school case studies do not teach. The same company, under different market conditions, has a completely different likelihood of becoming a market maker. Timing and context matter. There are certain questions you don’t know to ask until you’ve learned those lessons the hard way.
It is with that lens that I lay out my diligence process.
Why This Moment Is Different
The post-Cold War assumption of a stable, rules-based international order is over. The United States is managing simultaneous strategic competition with China, active conflict across multiple theaters, and a domestic political environment that has reshaped the relationship between government and private capital more dramatically than any period since the 1980s. Defense spending is rising. Middle Eastern and European sovereign wealth funds are deploying at unprecedented scale. National industrial policy has returned to Washington with a conviction that has not existed in decades.
Economically, capital has concentrated. The largest rounds are larger than ever. The median round is flat. The investors who win will identify structural inevitability before the market does.
Technologically, foundation models have commoditized general-purpose AI. Having an AI product is no longer a competitive advantage. Having a proprietary data asset, a physically embedded workflow, or a regulatory position that took years to earn is. The market has not internalized this distinction. Many companies valued at AI premiums are running on borrowed time.
The Filters
Filter One: The Right Market at the Right Time
The size of the problem has to justify the ambition of the company. The serviceable addressable market has to be large enough that winning a meaningful share produces a company worth building, and credible enough that the demand is structural and durable, not cyclical or dependent on a regulatory moment that may not last. The markets I find most compelling are the ones where the underlying need is non-negotiable. People do not stop needing medical devices because the economy slows. Militaries do not stop needing precision capabilities because procurement cycles are long.
But size alone is not enough. Being right about a market at the wrong time is expensive- something I learned the hard way, through my company Evalise, which was 15+ years too early for its time – doing voice based tech before Siri, Alexa and other devices came out. I look for the specific combination of structural inevitability – the forces, technological, geopolitical, economic, or regulatory, that make the company’s success not just possible but required – and timing that confirms the inflection is now. The clearest signal is that early customers are not buying because they are adventurous. They are buying because the alternative has become untenable. When bipartisan policy identifies a structural failure, when a technology cost curve crosses a commercial viability line, when regulatory pressure reaches a compliance threshold – those are not trends. Those are doors that open once.
The most durable version of this timing advantage converts into infrastructural indoctrination: technology woven so completely into the critical operations of its customers that removing it would be organizationally disruptive in ways no rational buyer would willingly accept. Systems integration depth, where replacing the vendor requires rebuilding adjacent infrastructure. Mission criticality, where the customer’s primary function now depends on the company’s output. Institutional knowledge transfer, where the company has accumulated an understanding of the customer that no competitor could replicate without equivalent years of exposure. Every additional month of deployment deepens the integration and raises the cost of replacement. The company that achieves this is not competing for renewal. It is being renewed by default.
What does not pass this filter: companies that are valuable but optional – optional gets cut – and any healthcare thesis that requires the U.S. health system’s incentive structure to change before the business can scale. The system responds to its actual incentives, not its stated ones. Betting on a system transformation that has not happened in two decades is not a thesis. It is a wish.
Filter Two: Defensible Competitive Advantage
Every founder believes their advantage is unique. Very few are right. My test: if the company raised no new capital for three years, would the competitive advantage be larger or smaller at the end of that period? If smaller, it is not a moat. It is a head start.
I look for replication resistance grounded in physics or biology – a core mechanism that cannot be shortcut regardless of capital. No amount of money compresses twenty years of clinical validation. No recruiting budget acquires a decade of physical-world operational data. I look for businesses where the technology becomes harder to challenge with time: compounding data flywheels, physical manufacturing infrastructure, regulatory track records that took years to build, network effects that deepen with each deployment. The strongest defensibility combines formal IP protection with operational depth that makes the technology more valuable with every year it stays deployed.
What does not pass: a software layer on top of a foundation model with no data moat, distribution advantage, or regulatory position beneath it – the model will commoditize and the layer will follow – and any company whose technology lead is replicable within five years by a well-resourced competitor. The businesses worth watching are the ones where the gap widens with time, not narrows. The more specifically I can answer why this company wins against each named competitor – mechanistically, not vaguely – the more confident I am that the advantage is real.
Filter Three: 10X Value Proposition and Clear ROI
Better is not enough. Measurably, demonstrably, irreversibly better – in a way the customer can feel and quantify – is a different class of investment. I use a simple test: is this product at least ten times better than the best available alternative along the dimension that matters most to the buyer? Not incrementally improved. Categorically superior.
The ten times threshold is not arbitrary. It is the minimum gap required to overcome institutional inertia, procurement friction, training costs, and the natural human preference for the familiar. Below that threshold, the better product often loses anyway.
Closely related is whether the value of implementation to the customer clearly outweighs the cost. In the categories I invest in, that calculation is complicated by switching costs, integration risk, training requirements, and organizational disruption. A product can be genuinely superior and still fail commercially because adoption costs more than the customer believes they will get back. The best investments are not sold. They are procured because the buyer’s own financial analysis demands it – in lives saved, cost avoided, revenue generated, or risk eliminated.
What does not pass: companies where the buyer cannot construct the ROI case without being coached through it – the product is not ready for market – and companies that depend on voluntary behavior change without a financial or structural driver. That model has failed consistently across health, finance, and consumer technology. The commercially rational decision and the structurally beneficial one have to be the same decision.
Filter Four: Commercial Architecture and Repeatable Revenue
Technology without a credible buyer is a science project. This is where the most optimistic founders fail – not on the technology, but on the commercial reality of how decisions actually get made inside the organizations they are selling to.
I want to know exactly who signs the check, who influences the decision, what the approval process looks like, how long the sales cycle runs, and what it costs to acquire and retain a customer. I have lived this lesson directly. At Asha AI, we built a product the pharmaceutical industry needed. We did not understand the procurement architecture – the layers of approval, the compliance requirements, the budget cycle timing, the internal champion dynamics – until we were already deep in sales conversations that were not progressing. The pharma funnel killed us. That experience taught me to treat commercial architecture with the same rigor as the product roadmap.
Equally important is what happens after the first sale. A business that generates revenue once is not the same as a business that generates revenue reliably. I look for multi-year contracts, subscription models with real retention, and customer relationships that expand in scope rather than expire at renewal. If the customer has not experienced genuine indoctrination into the product by the second renewal conversation, the moat was never there to begin with.
What does not pass: companies that have solved the technology problem but not the distribution problem – in healthcare, defense, and space, the buyer is risk-averse, compliance-driven, and relationship-dependent, and navigating that buyer is as important as building the technology – and revenue concentrated in a single government contract. Program cancellations are real, and concentration at that level is a valuation fiction.
Filter Five: Navigable Regulatory Landscape
I would not back companies whose primary risk is regulatory destruction of their business model. A single reclassification or reimbursement change can be existential. There is no operational solution to that problem.
I look for one of three positions. Regulatory clearance as a structural moat: an FDA De Novo approval, FAA license, or DoD program of record that took years to earn and cannot be shortcut by capital. Regulatory neutrality: a business operating in a stable zone where no single policy change disrupts the model. Or regulatory tailwind: government urgency actively creating demand for what the company does. When bipartisan need identifies a structural failure, the companies already positioned to solve it receive contracts. The ones still building do not. The companies I find most compelling have done the regulatory work upfront – not as a compliance exercise, but as a competitive strategy.
Filter Six: Unit Economics That Scale
A business that works at one customer does not automatically work at one thousand. Gross margins above 60 percent in software and above 40 percent in hardware-intensive businesses. Customer acquisition costs recovered within a year or two of the relationship. Marginal cost structures that improve with scale – software embedded in hardware already deployed, data products more valuable as the dataset grows, services cheaper to deliver as the team builds domain expertise.
What does not pass: businesses that require structurally high COGS that cannot be engineered down over time, and customer acquisition models that consume an unsustainable fraction of the revenue they generate. If the unit economics do not improve as the business scales, it does not get stronger as it grows – it gets more exposed.
Filter Seven: Founders Married to the Problem
The quality of judgment required to evaluate founders cannot be acquired by sitting across conference tables from entrepreneurs while never having been one. It is built through the specific texture of having been that person – carrying payroll, navigating a pivot under investor pressure, making consequential decisions at 2am with incomplete information and no one to escalate to.
What I look for is founders who are married to the problem, not to their current idea. A founder married to an idea defends it when the evidence turns. A founder married to the problem adjusts the idea when the evidence demands it. The best founders are simultaneously obsessed with their destination and genuinely indifferent to being wrong about their current route.
The other qualities cannot be faked across multiple conversations: intellectual curiosity that applies to every dimension of the business. The ability to communicate the thesis clearly without oversimplifying it. Competitive energy directed at the problem. Creative problem-solving that does not stop at the first-order answer. Emotional groundedness to know when to hold and when to pivot. These qualities are invisible in a pitch deck. They live in how a founder describes their last major failure, and whether they learned from it or rationalized past it. In the quality of their questions when pushed back on. In the gap between what they say their constraints are and what they are actually doing about them.
The Macro Conditions Shaping My Thesis
The American industrial reckoning. The U.S. dismantled its manufacturing base over forty years and is now confronting the national security consequences in real time. The defense industrial base cannot produce precision components, autonomous systems, or advanced platforms at the speed and volume modern deterrence requires. Companies positioned to address these gaps are not chasing a trend. They are solving a problem with no alternative solution.
The real structure of healthcare economics. Value-based care has been the stated priority of every administration for two decades. It has not arrived at commercial scale, and is unlikely to in the near term. The U.S. healthcare system runs on fee-for-service billing, CPT codes, hospital capital budgets, payor-by-payor coverage decisions, and pharmaceutical economics built around patent life and specialty drug pricing. The companies worth watching are not betting on a system transformation that has not happened. They are solving real problems inside the system that actually exists. Medical devices with FDA clearance create their own reimbursement pathway. Employer benefit spending is a validated channel that does not require payor reform. The commercial logic is already in place for those who know where to look.
The capital geography shift. The largest sovereign wealth funds in the Gulf are not chasing quarterly returns. They are building 20-year positions in technologies that will define national economic competitiveness. Understanding what that audience values, and how they evaluate defensibility, is itself a competitive advantage.
What I Know After Nearly Two Decades of Building
I have spent nearly two decades making bets on the idea that technology at the intersection of biology, intelligence, and physical systems would be the most consequential category of the next generation. That conviction was not fashionable when I started building. It is becoming obvious now.
What I look for in every company is the irreversible combination of a problem that cannot wait, a market large enough to matter, timing that is now, technology that gets harder to challenge with every passing year, a value proposition that is categorical rather than incremental, a commercial funnel that is understood and executable, revenue that compounds rather than resets, and a presence inside their customers’ operations that makes replacement unthinkable. When all of those conditions are present simultaneously, the outcome is not a function of market conditions. It is a function of execution.
Execution is the one thing I know how to evaluate, because I have done it myself – across multiple companies, multiple continents, and industry cycles that looked nothing like the one before them. I have navigated enterprise sales processes designed to outlast vendor patience. Rebuilt product architecture mid-flight when the original thesis proved wrong. Raised capital in markets actively hostile to the category I was building. Managed the tension of being too early, when the team believes and the market has not yet arrived.
The founders I back get access to that. I have already made most of the mistakes they are about to make and have built the instincts to recognize before they happen again, and the experience to know how to navigate many of them.
The companies featured across this series represent my personal watchlist and research interest. I have not personally invested in any of them, and nothing written here constitutes investment advice. The views expressed are my own.
