The Twenty Minute VCAdarsh Hiremath @ Mercor: The Fastest Growing Startup in Silicon Valley | E1261
CHAPTERS
- 0:00 – 0:34
$100M raise at $2B: framing Mercor’s scale and ambition
The episode opens with the headline funding details—$100M raised at a $2B valuation—then tees up Mercor’s broader thesis: many specialized AI models, recruiting as leverage, and network effects as software commoditizes. These themes become the backbone for the conversation that follows.
- •$100M round priced at a $2B valuation
- •Preview of Mercor’s worldview: many models, specialized use cases
- •Recruiting/talent flow as a company’s most important control point
- •If software approaches zero cost, network effects become the moat
- 0:34 – 2:32
Debate champions to co-founders: why debating maps to startup building
Adarsh explains how years of competitive debate with Surya (and later Brendan) shaped their founder dynamics. He draws parallels between debate partnerships and founding: high ownership, tight feedback loops, and the outsized importance of choosing the right partner.
- •Early relationship: meeting Surya at age 10, competing in debate together
- •Policy debate partnership as “first startup” with shared outcomes
- •Fast feedback loops and post-mortems mirror startup iteration
- •Founding team selection as the highest-leverage decision
- 2:32 – 3:34
From dev shop to labor marketplace: discovering the product by recruiting great people
Mercor started as a practical dev shop, but the founders noticed the real unlock wasn’t code—it was finding exceptional talent (initially in India). That insight drove them to automate candidate discovery and then automate the company-facing side, creating a two-sided marketplace.
- •Started building software for startups as a dev shop
- •Recruiting standout talent (especially from India) became the real advantage
- •Automation begins on the candidate side, then expands to the company side
- •Marketplace emerges as scaling requires automating sales/placement workflows
- 3:34 – 6:22
Harvard dropout decision: making the call before it was rational
Adarsh describes being physically at Harvard but mentally focused on the company, including a classic dorm-room founder story. Dropping out wasn’t obviously correct—no major funding, limited traction—but it became an emotional commitment to building with close friends.
- •Dorm-room hustle and early team dynamics (including future teammate Artemis)
- •Parents’ skepticism: no seed/Series A yet, limited revenue at the time
- •Decision driven by desire to work with best friends, not a checklist
- •Advice: founders often rationalize, but the choice is usually emotional
- 6:22 – 7:39
Seed round reality: $500/month salaries, fast close, and early moves
The conversation moves into their first capital raise and the visceral moments that made it feel “real,” like slashing salaries to $500/month. Adarsh recounts raising just over $3M quickly, led by General Catalyst, and learning from an early location decision (NYC vs SV).
- •Moved to New York pre-seed; later viewed as the wrong call
- •Symbolic milestone: setting founder salaries to $500/month in Gusto
- •Seed round: raised $3M+ and closed quickly
- •General Catalyst led; early investor relationships formed fast
- 7:39 – 10:37
Hypergrowth operating system: 9/9/6, ambition density, and what breaks first
Harry presses on Mercor’s rapid scaling and intense work cadence. Adarsh frames 9/9/6 as a byproduct of mission-driven hiring and momentum, while noting that extreme growth functions as a continuous stress test that forces constant role expansion and process rebuilding.
- •9/9/6 framed as avoiding Sunday work, not glorifying grind
- •Selecting for “care” as the hardest-to-teach trait
- •50% MoM growth as a perpetual stress test: process, hiring, roles break first
- •People must ‘outgrow themselves’ as the company’s needs rapidly evolve
- 10:37 – 11:09
Scaling culture faster than software: preserving the early core as headcount grows
Adarsh argues culture is harder to scale than product. The culture established in the first ~20 people is often the strongest it will ever be, so leadership must deliberately preserve it while hiring quickly and expanding into new workstreams.
- •Culture degrades easily during rapid hiring and organizational change
- •Early team sets the cultural “ceiling” unless reinforced intentionally
- •Maintaining intensity while keeping it sustainable is non-trivial
- •Legendary companies are built as much on cultural continuity as product
- 11:09 – 12:52
Mercor’s wedge: human data work becomes talent assessment for AI labs
Harry challenges whether Mercor is “just data labeling.” Adarsh reframes: modern post-training work requires domain experts, and identifying them is fundamentally a talent assessment problem—directly aligned with Mercor’s mission of building a unified labor market.
- •Old labeling: crowd tasks (e.g., bounding boxes) vs today’s expert-driven work
- •Post-training requires experts to improve models in specific domains
- •Finding the right expert is the core product: assessment + matching
- •AI lab needs mirror Mercor’s long-term goals (talent pool + performance prediction)
- 12:52 – 24:08
How the product works: AI interviewer, automated pipeline, and retention as the north star
Adarsh details how Mercor places talent—including for top AI labs—using an automated end-to-end workflow. He emphasizes net retention as the core success metric and describes the “AI interviewer” as generalizable infrastructure that can generate role-specific interviews in seconds.
- •AI labs hire via Mercor similarly to traditional roles; often for post-training
- •Core metric: customers expand usage; net retention is well above 100%
- •Generalizable AI interviewer creates custom interviews in ~10 seconds
- •Mercor automates the full candidate flow from intake to interview to payment
- 24:08 – 27:01
Quality over price: finding the top 0.1% and winning the sales moment without sales
Mercor’s sales ‘wow’ happens when the first candidates start delivering value. Adarsh explains why quality is the dominant buying driver (unlike commoditized gig work) and notes Mercor runs with no traditional sales team, relying heavily on inbound and word-of-mouth.
- •Customer wow moment: first placements performing strongly
- •Quality gap is nonlinear (top 0.1% vs 80th percentile)
- •No dedicated sales team; founders handle relationships; inbound-heavy growth
- •Take rate varies case-by-case; price is secondary when quality is exceptional
- 27:01 – 28:23
Marketplace expansion: from India roots to majority-US workforce and strong network effects
Adarsh explains why Mercor initially focused on India (founder connection and recruiting advantage), then how the platform expanded so the largest worker base is now in the U.S. He also breaks down Mercor’s two moats: two-sided marketplace liquidity and a job-performance data flywheel.
- •India starting point: targeted school campaigns; early ‘manual’ recruiting lessons
- •Today, majority of placed workers are from the United States
- •Two network effects: marketplace liquidity + performance prediction flywheel
- •Outcome data enables better matching—even when candidates don’t know best-fit roles
- 28:23 – 30:36
AI and the future of programming: abstraction shift, agentic tooling, and software commoditization
Harry asks whether young people should still study programming as AI writes more code. Adarsh argues programming becomes more important but moves up the abstraction stack—humans orchestrate fleets of coding agents—and tools like Cursor accelerate testing, refactoring, and consistency work.
- •Programming persists but shifts from code-level to orchestration-level control
- •Analogy: Assembly→Python was as big as Python→natural language
- •Tools like Cursor change workflows: faster tests, refactors, consistency
- •Resulting macro-effect: software creation cost collapses and commoditizes
- 30:36 – 35:38
Who wins when software is ‘free’: SaaS becomes services replacement, and moats are network effects
Adarsh outlines a world where shipping software is easy and cheap, making code alone non-defensible. He argues enduring businesses will be built on network effects, and “next-gen SaaS” will replace entire end-to-end services rather than sell feature-based tools.
- •Software commoditization shifts advantage away from proprietary code
- •Network effects become the core moat (marketplaces like Airbnb/Uber/Meta)
- •SaaS evolves into end-to-end service replacement (Mercor as example)
- •Stickiness comes from quality and usage-based expansion, akin to Stripe
- 35:38 – 40:48
Fundraising stories: Benchmark helicopter, fast round cadence, and the $100M balance-sheet logic
The discussion turns to Mercor’s rapid fundraising cadence and why they took money despite strong revenue momentum. Adarsh shares the memorable Benchmark ‘helicopter’ moment, explains they don’t enjoy fundraising, and frames the $100M raise as long-term balance-sheet strength for a labor aggregation mission.
- •Benchmark origin story: Victor Lazarte intro; helicopter ride with Peter Fenton
- •Rounds often ‘came to them’ while they stayed heads-down on execution
- •Board composition: founders + Benchmark
- •$100M round led by Felicis; goal is resilience for a long, ambitious build (not immediate spend)
- 40:48 – 46:23
Quick-fire philosophy: recruiting prestige, personal discipline, AGI curiosity, and 2035 vision
In the closing quick-fire, Adarsh shares contrarian beliefs and personal reflections, then paints a big-picture future for Mercor. He argues recruiting is the highest-leverage function, reiterates SaaS-as-service replacement, and imagines Mercor becoming the default global labor market by 2035.
- •Contrarian belief: recruiting is the highest-prestige, most leverageful role
- •Efficiency depends on matching the right person; manual processes don’t scale
- •Personal reflections: pushing like an athlete; what to stop; what he’d known earlier
- •2035 vision: Mercor as unified labor marketplace facilitating massive job matching at global scale
