Why India’s Next Tech Giants Are Hiding in Plain Sight | Paavan Gami
Topics
Early-Stage VC Investing Dynamics
AI Workforce Displacement and Labor Disruption
AI Data Center Power Bottlenecks
Crypto Trading Psychology and Risk Discipline
Pioneers
Paavan Gami
Summary
- India’s early IPOs let Gami pursue venture-style compounding in public markets. Macedo frames Greenoaks with its roughly 40% first-fund investment in Coupang, which he says returned about $8 billion. Gami cites a roughly two-year-old median IPO valuation of $350 million, with founders or families retaining 70%; category-leading technology companies can list at $600 million–$1 billion. He sees an opportunity to buy misunderstood, reinvesting businesses at an estimated 15–20 times long-term earnings: the American pattern of staying private longer is “the opposite” of what happens in India.
- The stock-picking thesis rests on market-share gains as much as India’s economic growth. Gami sees leaders with 5–6% share potentially reaching 30–40% as fragmented industries consolidate and informal operators lose their compliance advantages. Macedo points to an e-commerce grocer he owns with 2% share, whose growth he thinks remains largely controllable whether GDP grows 7% or 4%, despite India’s institutional “ankle weights.”
- Eternal is Gami’s bet that an Indian delivery leader can become a technology conglomerate. Blinkit reached $10–11 billion in top line in roughly four years, while mature locations offer 80,000 SKUs within 15 minutes. Ras entered at 1.3–1.4 times forward GMV; assuming eventual 5% free-cash-flow margins implied roughly 30 times free cash flow for a business then growing 70%, with its founder acting as “head of HR.”
- Meesho’s investment case combines low prices with logistics density that competitors would struggle to replicate. Gami claims 70–80% of e-commerce in tier-three and tier-four cities, with roughly $3 orders against competitors’ $15–20. Its operating priority is to “lower prices every single year”; financial services could become a next act for its 300 million largely unbanked customers.
- The AI discussion weighs productivity gains against risks to India’s labor and outsourcing advantage. Macedo raises possible education and junior-worker productivity gains; Gami argues India benefits so long as labor remains useful, while acknowledging a no-labor future would be bad for the country. Gami is more interested in businesses that benefit from AGI than in businesses merely insulated from it, and the discussion remains concerned about employment even if IT-services companies survive.
- India’s data-center opportunity could extend beyond domestic demand into global inference and equipment supply. Gami cites build costs of $6 million per megawatt against $16 million in the US, with latency the constraint for some workloads. India could reach 8–10 GW by 2030 on domestic demand, or 20 GW if it serves global demand, while suppliers targeting “the COGS of Schneider” offer another route—he is building an unnamed small-cap position.
- Carvana taught Gami that surviving the drawdown and finding the best entry are different underwriting problems. Greenoaks rebuilt a position after a 90% decline, then endured another roughly 75% fall. He thinks the later $30–40 price after debt restructuring probably offered the best risk-reward, when most capital was already committed: “could we have saved a bullet?”
- Ras’s concentration depends on accumulated founder knowledge, with AI increasingly handling monitoring. Larger positions run 10–12% at cost, and building the surrounding research network takes six–12 months. Agents can flag changing thesis conditions, but Gami still values disagreement and human judgment; his biggest identified risk is “I don’t execute well.”
Deep dive
1. Greenoaks taught Gami to design the firm around conviction
- Macedo discloses that he is an LP in Ras Partners and introduces Gami as a five-year Greenoaks alumnus. Macedo frames Greenoaks with its roughly 40% investment of its first fund in Coupang, a bet he says returned about $8 billion. Holding winners after their IPOs, rather than treating the listing as an exit, is central to the approach he describes.
- Investing clicked for Gami through the Nomad letters: people and cultures became “the machines that make the machine,” moving him beyond arbitrage and buying 50 cents on the dollar. As SPO Partners’ first internet analyst, researching Naver’s shopping ambitions introduced him to Coupang investors Neil and Benny. He joined Greenoaks at the end of 2018.
- Gami describes Greenoaks as a system with a “single decision-maker,” few investments and flexibility across early-stage, late-stage, public and structured opportunities. His takeaway for Ras was to build around his own strengths: opportunity selection, research process and eventually organization design. He wanted a firm that would let him “make a 20-year bet on something.”
2. Carvana required underwriting survival before underwriting scale
- Gami liked Carvana’s solution to the used-car “lemon problem,” alongside scale economies in refurbishment, parts and logistics. Greenoaks owned it from mid-2020 through the end of 2021, selling roughly 20–30% below the peak on valuation. Re-entering in the $20s after a 90% decline preceded another approximately 75% fall over six–eight months. Knowing the business mattered: “it’s very, very hard to run into a burning building.”
- The crisis combined weaker demand, rate-sensitive customer financing, higher inventory-financing costs and a large leveraged acquisition. Gami recalls quarterly cash burn reaching $150–200 million, with three–four quarters of runway. Two quarters without substantial cost cuts raised the possibility that expenses were structurally inflexible. His counterexample was March–May 2020, when Carvana had cut fixed costs sharply; creditors’ incentives and restructuring options required separate work.
- The deeper question was whether demand survived sustainable economics. Carvana appeared to sell cars at market prices, rather than manufacture growth through discounts; the remaining challenge was establishing a reasonable cost to serve. Gami also challenged the subprime narrative through ABS filings: competitors sold subprime loans wholesale while Carvana retained them, making its reported mix look worse. He spent most of that year on these questions beyond private-company board commitments.
- Gami’s compressed thesis was an exceptional entrepreneur and category-killing retailer at roughly 5× equity to free cash flow: “if they weren’t bankrupt, you were going to do well.” He thinks the best risk-reward probably arrived after renegotiation converted interest expense into PIK, with shares at $30–40 and bankruptcy risk substantially reduced. Most capital was already deployed. His unresolved lesson: “could we have saved a bullet?” Decisions were being made under uncertainty.
3. India offers consolidation alongside economic growth
- Gami’s macro case starts with 1.5 billion people, half under 30: roughly 20% of humanity but 4–5% of global GDP. Real GDP has grown 6–7% annually for 25 years despite institutional shortcomings and what he considers populist policy mistakes—“GDP growth with ankle weights on.” Liberalization began in 1991; he sees continued incremental reform rather than an already perfected system.
- Across pharmacies, grocery, hotels, hospitals and other consumer categories, Gami estimates roughly 90% of supply remains independent. Historically, “regulatory diseconomies of scale” protected small operators: large companies had to pay employees formally, incur social-security costs and transact compliantly. As smaller businesses are also pushed toward compliance, the disadvantage of being large diminishes and procurement, operating and distribution advantages become more valuable.
- The investment is a leader moving from 5–6% share toward 30–40% over 20–30 years while its category expands; Home Depot’s roughly 40% US share supplies a template. Macedo adds a portfolio example: an e-commerce grocer he owns has 2% share, and he thinks its growth is largely controllable whether Indian GDP grows 7% or 4%.
4. AI challenges the demographic thesis without settling it
- Macedo raises the possibility that AI could turn India’s young population from an advantage into a burden, and frames a post-scarcity scenario in which countries with less capital and fewer fixed assets could be left behind. The exchange also includes the view that AI has made India less attractive. Gami’s explicit caveat is that a world with no labor in the production function would be bad for India, though he calls that a bizarre limit to contemplate.
- Macedo argues that weak education could leave substantial room for AI-assisted learning and that junior Indian developers might gain more revenue per head than already exceptional American engineers. Gami’s response is that, so long as labor remains useful, higher productivity and upskilling could benefit India; replacing an inexpensive Indian driver with an autonomous vehicle is also a different break-even calculation.
- The IT-services discussion raises both a possible survival path and an employment risk. Enterprise-specific workflows and proprietary data could shift budgets from Oracle licenses toward building custom ERP systems, with tokens becoming a major delivery cost. That could benefit IT-services companies while still threatening employment; the speaker discussing the sector says he is “reasonably worried” about that outcome.
5. AI has changed which Indian businesses deserve attention
- The discussion revisits traditional businesses featured in Ras’s launch letter two years earlier. The speaker argues that avoiding disruption is insufficient if a business’s future is not accelerating through AI and internally built technology; the opportunity cost of investing in companies that are accelerating has risen. The discussion also describes learning industrial businesses despite constitutional skepticism of bottleneck stocks, accepting that moving slowly may mean missing opportunities.
- Culture can override the business-model label. Gami asks listeners to imagine a 20-year-old Walmart still run by Sam Walton: he thinks Walton would embrace AI aggressively. India’s young companies, often led by entrepreneurs aged 40–60, might adapt faster than mature Western counterparts. Some ostensibly traditional businesses also emerged from venture-backed ecosystems and already possess technology-oriented cultures.
- Macedo relays Grad Capital founder Abhishek’s concern that India may have “maybe one year left” to produce generational technology companies before frontier talent concentrates in San Francisco. Gami agrees that frontier-adjacent AI businesses requiring hundreds of millions of dollars are difficult to finance locally. He also sees relatively few exciting private companies worth $1–3 billion.
- Gami disputes the broader pessimism about early-stage activity. He sees energy in defense, aerospace and biotech, with some founders motivated to build for India. Those companies, he argues, may possess more terminal value than another Lovable competitor. Meanwhile, existing public technology businesses have distribution, cash-generating cores and talent that could let them capture new markets, including Indian AI applications.
6. Small IPOs and unrepresentative indices create a different public market
- Gami cites a roughly two-year-old statistic: the median Indian IPO market capitalization was $350 million, with founders or founding families retaining 70%. Underdeveloped private capital markets mean companies often bootstrap, use debt for factories and sell their first equity at listing. He sees the continuing family ownership as a source of alignment and capital-efficient execution.
- Technology companies seeking $100–200 million may find an IPO easier than a Series D. Category leaders can list at $600 million–$1 billion, leaving subsequent compounding accessible to public investors. Negative margins and heavy reinvestment complicate conventional valuation; Gami thinks some durable businesses are available at 15–20 times long-term earnings. His claim concerns future earnings power, rather than reported profitability today.
- The index contains public-sector businesses, cyclicals and low-return-on-capital industries, making it a poor proxy for his opportunity set. Gami counts one technology company among the top 50 market caps, around number 40, and approximately three among the top 100. He expects high-rate compounding over five–seven years, with index inclusion and liquidity flows possible later benefits rather than the investment rationale.
7. Founder underwriting determines which holdings earn lasting commitment
- Gami’s central evolution at Greenoaks was internalizing the “night and day” difference between exceptional and merely great entrepreneurs—and treating that distinction as researchable. Pattern recognition is supplemented by extensive references and time with founders and their direct reports. The goal is to understand how the organization actually works and become a useful thought partner.
- Ras has “married two businesses” in its first two years; Gami thinks finding one or two annually might be realistic. Marriage means enough conviction in the founder and culture to give the business the benefit of the doubt and buy more when trouble arrives. Below those highest-conviction holdings, he describes continuing “creative destruction” as ideas compete for capital.
- With roughly 10 holdings, Gami describes larger positions at 10–12% of cost and smaller starters. Each requires substantial “upfront CapEx”: building relationships around the company over six–12 months, including senior former employees. Understanding whether a product-led organization can execute an operations-heavy initiative requires that accumulated context. Starter positions grow as conviction develops through the work.
8. Eternal combines delivery economics with a repeatable talent system
- Gami traces Eternal, formerly Zomato, from restaurant classifieds across 25 countries into food delivery. It entered roughly a year after competitors but won with less capital than Swiggy. He describes today’s market as a 60/40 duopoly, with both earning above 5% margins, and calls it the world’s most profitable food-delivery category. Its conglomerate ambition resembles Tencent: deploy cash and distribution into additional businesses.
- Blinkit reached $10–11 billion in top line within roughly three-and-a-half–four years. Gami says it became India’s second-largest grocer within three years and will probably become the largest within 12 months. Mature markets offer 80,000 SKUs in 15 minutes, versus roughly 15,000 at overseas quick-commerce peers. Weak incumbent grocery chains gave it a procurement opening that an upstart competing against Aldi would lack.
- Adoption reflects “compensating for the lack of urban infrastructure”: traffic and parking make rapid delivery useful, while prices match or beat neighborhood shops. In mature geographies, roughly a quarter of households use it. The founder’s “head of HR” role moves talent between businesses; after the 2021 IPO, Gami says 10 of 15 senior leaders were pushed out for complacency and replaced by younger employees.
- Ras built its position roughly 18 months earlier at 1.3–1.4× forward GMV. Assuming eventual 5% free-cash-flow margins across business units implied approximately 30× free cash flow while growth was 70%. Offline retailers traded around 3× GMV and 60–70× earnings. Gami saw faster growth and prospective dominance available at roughly half the offline valuation multiple.
9. Meesho trades delivery speed for prices and distribution reach
- Meesho is Gami’s PDD- or Shopee-like value-commerce investment: breadth and low absolute prices take precedence over speed. He thinks this format is likely larger than Amazon-style commerce in lower-income markets. It began enabling WhatsApp resellers; when end customers started buying directly at attractive wholesale prices, the business evolved into a marketplace.
- Gami calls it a monopoly, claiming roughly 70–80% of e-commerce in tier-three and tier-four cities, where he says 1.2 billion of India’s 1.5 billion people live. Average orders are approximately ₹250, or $3, against competitors’ $15–20—roughly the larger players’ logistics cost alone. Meesho does not own the trucks, but its volume creates utilization advantages that new entrants would struggle to match.
- The operating priority is to “lower prices every single year,” through seller competition and logistics efficiency. To bridge the gap between Bangalore-based IIT engineers and small-town customers, every employee spends their first two weeks interning with a seller and must deliver business value. Leadership spends a week each quarter with customers, embedding that understanding into product and operating decisions.
- Gami thinks the “next act” will likely be financial services. Meesho has 300 million customers who are largely unbanked and need credit; he suggests perhaps beginning with BNPL and eventually moving into unsecured consumer lending. This remains a prospective extension of the platform, rather than an established earnings stream in his account.
10. Data centers offer domestic demand and a possible export supply chain
- Gami began researching the opportunity six–eight months earlier and regrets not starting 18 months ago. Government and financial-sector localization demand supplied the initial rationale. He then considered global inference: latency constrains some workloads, but long-horizon tasks could run in India. He cites construction costs of $6 million per megawatt versus $16 million in the US, alongside inexpensive land, power and skilled labor.
- India starts with approximately 1.2–1.5 GW of data-center capacity, little of it GPUs. Additions are around 300 MW this year; announced projects imply 1.5 GW annually in three years, a fivefold increase. That could produce 8–10 GW by 2030, consistent with India-for-India demand. If India serves substantial global demand, Gami suggests 20 GW instead of 10 GW.
- Most spending would leave India through compute and memory purchases, but approximately 30% of the bill of materials involves HVAC, piping and cooling that could localize. Suppliers to Schneider and ABB might then win global business. Ras is building a position in an unnamed company worth a few hundred million dollars that Gami says is supplying Middle Eastern and European build-outs; he also expects some planned Middle Eastern capacity to shift toward India.
11. AI is becoming a monitoring system around human judgment
- Gami wants company context in his own head so he can update quickly when facts change; simply reading an AI-written investment memo does not solve that. His bandwidth allows deep work on perhaps a couple of companies monthly. The useful problem is recognizing when idea number 18 on the watchlist deserves attention immediately, rather than months later.
- His practical application is a “fancy alerting system.” After researching and passing on a company, he records conditions that would change the decision. An agent reads news and earnings calls monthly, looking for fundamental inflections: “The answer was no. Now maybe it’s yes. You should revisit it.” That is more useful to him than another generic equity-research report.
- His richest insights still come from people, including private competitors trying to recruit portfolio-company executives or attack their moats. Recruitment difficulty and operating anecdotes test the investment thesis. Autonomous expert calls and customer surveys seem technically possible, but his experiments have been weak enough that he prefers doing calls himself today. He wants more desk work automated to preserve time for those relationships.
- AI has deferred a hire he otherwise thinks he would have made, but trusted judgment and communication remain valuable. A colleague who disagrees and argues for a week can change his mind; “the AIs are too sycophantic to do that.” Their “sloppy prose” also undermines his confidence. He still expects to hire people who can form views and persuasively defend them.
12. Ras’s ambition depends on execution and relationships built over years
- Asked for the biggest risk, Gami answers: “I don’t execute well.” A narrow strategy could create blinders that prevent him seeing where he should focus. India itself could disappoint, but he emphasizes controllable execution risks and keeps a list of structural and personal “potholes.” His national bull case centers on adaptable young entrepreneurs, including outside explicitly technological industries.
- Ras’s desired outcome is a relationship with one or two generational Indian founders over the next decade, supported by a high-quality capital base. His detailed letters serve that relationship: “the audience is the founder.” Exceptional entrepreneurs who remain hungry for 10–20 years, backed by profitable core businesses, could eventually need a capital partner as well as a thought partner.
- Gami uses Silver Lake’s Waymo investment to explain why a cash-rich parent might still seek outside capital: subsidiary equity can incentivize its own team. His ideal investor-founder relationship also permits funding a new idea within days. That speed comes from five–10 years of accumulated diligence, rather than everything learned during the subsequent 48 hours.