The Twenty Minute VCWe Built Our Own Salesforce in Months. Here's Why We're Cancelling the $600K Contract | Curative CEO
CHAPTERS
- 0:00 – 1:16
Curative’s COVID blitz: from no tests to 206k/day and 7→7,000 employees
Fred opens with the moment lockdowns began and testing was unavailable, while Curative’s CSO had already built a COVID test in spare time. The company rapidly scaled operations to peak at 206,000 tests in a single day and ballooned headcount from 7 to 7,000 in nine months.
- •COVID testing capacity was near-zero early on; demand spiked instantly
- •An internal COVID assay existed before the pivot became official
- •Peak throughput reached 206,000 tests/day
- •Hypergrowth in staffing: 7 to 7,000 employees in ~9 months
- •Operational urgency and speed defined the early pandemic period
- 1:16 – 1:55
What drives Fred: chasing the thrill of winning (and speed)
Harry frames founder motivation as fear of losing vs. thrill of winning. Fred describes becoming addicted to the momentum and pace Curative achieved during COVID and trying to recreate that velocity afterward.
- •Fred is motivated primarily by the thrill of winning
- •COVID created an unusually strong tailwind and organizational momentum
- •He values environments where speed is possible
- •Post-COVID, he kept chasing the same pace of execution
- 1:55 – 3:39
UK vs. Silicon Valley: why this company couldn’t be built in Britain
Fred explains why he believes the UK ecosystem—especially a decade ago—was less receptive to young founders without conventional credentials. He contrasts UK risk-mitigation behavior with Silicon Valley’s upside-optimization and willingness to bet on potential.
- •Difficulty getting investor meetings in the UK as an 18-year-old founder
- •UK investing felt credential-driven; Valley felt possibility-driven
- •Silicon Valley optimized for best-case outcomes vs. avoiding worst-case outcomes
- •Network effects: more repeat founders and angel/VC flywheel in the Valley
- 3:39 – 7:24
Origin story: cattle genetics startup → YC → A16Z seed (then TAM reality hits)
Fred’s first startup began by sequencing cows to predict traits like musculature and milk yield, sparked by a farmer’s outreach. After moving the company to the US and joining YC (S16), it raised a seed from Andreessen’s bio fund—only to later confront that the cattle testing TAM couldn’t support a big venture trajectory.
- •Started with cattle genetics testing in northern England
- •Applied to YC at the last minute after attending an AgTech conference
- •Moved to Silicon Valley for YC S16; raised ~$1.65M seed from A16Z Bio
- •Series A prospects forced a hard TAM calculation (cattle testing too small)
- •Set up the next pivot by reusing DNA-testing capabilities
- 7:24 – 10:02
Pivot into human diagnostics: high-throughput and at-home STD testing
Using the same underlying DNA testing approach, the team pivoted from animals to humans where willingness to pay and market size were larger. They built high-throughput STD testing and an at-home offering, focusing on antibiotic resistance and better drug selection.
- •Back-end tech continuity: DNA testing principles transfer across domains
- •Chose humans for larger markets and higher value per test
- •Focused on antibiotic resistance in STDs and better early drug targeting
- •Launched a high-throughput STD lab plus at-home STD testing
- 10:02 – 11:33
Sepsis: the high-stakes problem and why better tests alone weren’t enough
Fred explains sepsis pathophysiology and why time-to-treatment is critical (each hour untreated sharply increases mortality). He describes an insight: diagnostics work in academic centers but fail in community hospitals due to understaffing and slow suspicion—meaning adoption and workflow matter as much as test performance.
- •Sepsis kills at scale; immune overreaction drives organ failure
- •~12% mortality increase per hour without treatment (as described)
- •Better diagnostics succeed in academic centers but fail in broader trials
- •Community hospital constraints prevent early suspicion and test usage
- •Conclusion: workflow/operations can be the limiting factor, not science
- 11:33 – 14:51
Company near-death: the killed Series B, cash crunch, and the CLIA license timing lesson
The sepsis-focused company (renamed Shield) pursued FDA progress and a Series B led by a strategic—until the strategic’s CEO killed the deal as too competitive. With only weeks of cash, the company shut down; Fred sold a CLIA license for $150k, then months later had to spend $27M to regain equivalent capability during COVID.
- •Strategic term sheet collapsed late due to internal competition concerns
- •The company effectively died end of 2019 with ~3 weeks cash
- •CLIA lab license sold for $150k in wind-down to pay creditors
- •Five months later, a lab/license acquisition cost $27M during the pandemic
- •Key takeaway: timing and dependency risk can dominate outcomes
- 14:51 – 19:04
Founding Curative on a sepsis-services thesis—then COVID freezes hospitals
Curative originally aimed to operationalize sepsis care by embedding ‘mini-hospital’ capabilities inside community hospitals for a fixed fee. A pilot was lined up, but the partner hospital abruptly paused everything as COVID preparations began—signaling that the world was about to change.
- •Curative’s first pitch: take over sepsis management inside community hospitals
- •Business model: fixed-fee responsibility tied to outcomes
- •Seed round (~$1M) and early believers (Justin Mateen first check)
- •Wisconsin pilot paused because the hospital was preparing for COVID
- •Early clue that COVID would be bigger than expected
- 19:04 – 25:55
Overnight pivot to mass COVID testing: finding a lab, first customers, and a tweet that landed LA
With a test already built, Curative’s bottleneck was a licensed lab and partners willing to touch COVID. A JV with a small licensed lab in San Dimas enabled launch, and a tweet (via Laura Deming) led to a breakthrough conversation with LA leadership, turning pilot success into major contracts.
- •Mid-Feb realization from early spread data triggered action
- •Had a COVID test ready; lacked a lab/CLIA infrastructure to run it
- •No Bay Area facility wanted COVID on-site; found a licensed lab in San Dimas
- •Laura Deming’s tweet connected them to LA’s deputy mayor
- •LA paid extraordinarily fast (daily invoices; net-1 checks) to fuel scaling
- 25:55 – 27:06
Biggest deals and why Curative won: designing for 10× scale (Florida nursing homes)
Fred describes major contracts, including Florida’s statewide nursing home employee testing program worth hundreds of millions. He argues incumbent labs were built for incremental efficiency, not rapid capacity expansion, so Curative approached it as a greenfield scaling problem.
- •Largest contracts included Florida statewide nursing home testing
- •Competitors declined bids due to impossible volume and turnaround constraints
- •Incumbent labs optimize for margin efficiency, not 10× scaling
- •Curative bid aggressively and delivered by rebuilding assumptions from scratch
- 27:06 – 30:44
‘Orthogonal supply chain’: scaling through alternative inputs and manufacturing paths
To avoid the broken, crowded supply chain everyone chased, Curative redesigned inputs and processes to use less-contested materials and methods. Examples include alternative swab sourcing/sterilization and shifting away from magnetic beads toward filter plates to scale faster than constrained factories could support.
- •‘Orthogonal’ meant avoiding standard consumables everyone competed for
- •Used nontraditional swab supply and sterilization approaches
- •Avoided magnetic beads constrained to limited factories; used scalable filter plates
- •Partnered with vendors capable of scaling plastic/glass manufacturing
- •Result: rapid ramp to 206k tests/day within months
- 30:44 – 34:50
$5B revenue, surge economics, and why vaccines were a money-loser
Fred explains the economics: COVID testing was profitable during surges but loss-making during lulls due to fixed costs sized for peaks. Curative also administered 2.5M vaccinations but lost money on every dose because reimbursement didn’t cover administration cost—done largely to support partners and public need.
- •Total COVID-era revenue ~ $5B across ~3 years
- •Surges filled capacity and drove profits; dips underutilized fixed infrastructure
- •Reimbursement incentives encouraged overbuilding peak capacity
- •2.5M vaccinations delivered; government payments didn’t cover cost
- •~$500M ultimately funded the next pivot into insurance
- 34:50 – 40:33
Post-COVID ‘what next’: why Curative chose health insurance (and Fred’s critique of the system)
As COVID demand inevitably faded, Curative explored paths like lab testing, buying hospitals, and primary care—finding each constrained by TAM, payer mix, or incentives. They converged on insurance because the payer drives behavior in US healthcare, and Fred argues consolidation has broken efficient negotiation and inflated costs.
- •Started planning early because Fred believed COVID would be temporary
- •Rejected lab testing as too small even at industry-displacement scale
- •Considered hospitals but payer mix fragmentation limited transformation ability
- •Conclusion: payer side drives system incentives and behavior
- •System critique: consolidation creates negotiation deadlocks and inflated prices
- 40:33 – 45:45
AI agents inside an insurer: credentialing, claims, underwriting ingestion—and the ‘agent supervisor’ role
Fred details how LLM agents transformed back-office insurance workflows: credentialing dropped from months and ~$50 to ~12 hours and ~$0.20, and document/spreadsheet ingestion shifted to one-off Python generation. He predicts a new job category—agent supervision—because scaling agents creates an exception-handling bottleneck.
- •Credentialing agent verifies licenses/transcripts/malpractice checks end-to-end
- •Turnaround improved from 2–3 months to ~12 hours; cost from ~$50 to ~$0.20
- •Claims and underwriting workflows increasingly automated
- •LLMs used for codegen: generate one-off Python to normalize arbitrary inputs
- •Emerging role: ‘agent supervisor’ to manage exceptions and approvals at scale
- 45:45 – 1:13:27
Is SaaS dead? Cutting Salesforce, replacing insurance platforms, and slashing 80% of SaaS spend
Fred argues many SaaS tools will be replaced by internal, AI-built systems tailored to a company’s workflow. Curative canceled Salesforce ($600k/year) after building an internal CRM in two months, is migrating off expensive BI tools, and is even rebuilding core insurance systems whose APIs and data access were restrictive.
- •SaaS displacement driven by custom workflows + cheaper internal build via AI
- •Salesforce canceled ($600k/year); internal CRM built in ~2 months
- •Target: cut ~80% of SaaS spend; track renewals and non-renew decisions
- •Some infra SaaS may persist (e.g., monitoring); Slack sticky due to integrations
- •Rebuilt claims system in-house to escape vendor lock-in and poor APIs
- 1:13:27 – 1:29:35
Subcritical: a safer nuclear reactor via subcritical ‘energy amplifier’ + accelerator control
Fred describes co-founding Subcritical (with his wife) to push nuclear deployment faster in regulated environments. The core concept: operate below criticality (e.g., 0.97) and use a particle accelerator to supply extra neutrons—so turning off the accelerator stops the reaction, preventing runaway criticality and enabling a fundamentally safer design.
- •Nuclear is safe; biggest barrier is regulation and permitting, not physics
- •Traditional reactors balance near criticality; regulation demands extreme guarantees
- •Energy amplifier approach operates subcritical and uses an accelerator as an on/off switch
- •Design aims to make runaway criticality physically impossible under operating assumptions
- •AI also accelerates hardware design via code-generated CAD and optimization
