Science Corporation

Phase A — Understand the business

Lens 1 · Company Overview

Science Corporation (brand: "Science"; domain science.xyz) is a clinical-stage neuroengineering company founded in 2021 by Max Hodak — co-founder and former president of Neuralink, who departed Musk's company in May 2021. It is headquartered in Alameda, California, with ~150 employees. Hodak's framing is explicit and unusual for the BCI field: he wants a revenue-generating brain-interface business, reportedly targeting $100M of revenue, not an indefinite research project. That single sentence is the whole investment story — Science is the only serious BCI player monetizing now.

The business model in plain terms. Science is really four businesses stacked under one roof, sequenced from "ships this year" to "ships next decade":

1. PRIMA — the product. A wireless, photovoltaic subretinal implant (a chip "smaller than a grain of rice") paired with camera glasses and a pocket processor, restoring functional central vision to patients blinded by geographic atrophy (advanced dry age-related macular degeneration). This is the near-term commercial engine — CE-mark and FDA applications are in.
2. Science Eye — the next vision program. An optogenetic gene therapy (sensitizing surviving retinal ganglion cells) combined with an implanted ultra-dense microLED panel, aimed at retinitis pigmentosa and later dry AMD — still pre-clinical/animal-safety stage.
3. Biohybrid neural interface — the moonshot. Stem-cell-derived, genetically engineered neurons grown on a "waffle"-like scaffold that sits on the cortex and grows living axons into it — a fundamentally different bet from penetrating electrodes. Mouse-stage; preparing a first human sensor placement.
4. Vessel + Science Foundry — infrastructure/adjacencies. Vessel is a portable organ-perfusion / life-support platform (announced Dec 2025). The Science Foundry (Morrisville, North Carolina; acquired from MEMSCAP in late 2022) is a captive MEMS fab — vertical integration for its own implant manufacturing.

Customers / suppliers / competitors. End customers: retina surgeons and the patients they treat (initially in Germany/EU), reimbursed by national health systems and private payers — not consumers. Key suppliers: in large part itself (the captive MEMS Foundry for microfabrication), plus contract optics/electronics and AAV gene-therapy CDMOs for Science Eye. Competitors: Neuralink, Synchron, Precision Neuroscience, Paradromics (motor/communication BCI) on the moonshot side; on the product side PRIMA competes with anti-complement drugs for GA (Apellis' Syfovre/pegcetacoplan, Astellas/Iveric's Izervay/avacincaptad) — which only slow atrophy and do not restore vision, a crucial differentiation. Contract structure: none recurring yet — PRIMA will be a one-time implantable-device sale plus follow-on glasses/software, i.e. medtech economics, not SaaS.

Lens 2 · Supply Chain

Upstream → company → end customer, named where sourceable:

• Microfabrication (the implant chips): the Science Foundry in Morrisville, NC — the former MEMSCAP MEMS fab, acquired late 2022, expanded in 2024 with up to $65M of investment and 50+ jobs. This is the defining supply-chain move: Science owns its foundry, so the photovoltaic-pixel array and the biohybrid scaffold are made in-house rather than sourced from TSMC-class outsiders. [vertical integration — moat input]
• Optogenetics / gene therapy (Science Eye): AAV vector manufacturing — a standard biopharma chokepoint (CDMO capacity for clinical-grade viral vector is scarce and expensive). [chokepoint — outsourced]
• Stem-cell neuron supply (Biohybrid): in-house derivation of optogenetically modified cortical neurons from stem-cell lines — a wet-lab capability, not a purchased input.
• Glasses / camera / processor (PRIMA system): consumer-electronics-class optics + low-power compute — commoditized, multi-source, not a constraint.
• Originating IP: PRIMA's core photovoltaic-implant technology was invented at Stanford (Daniel Palanker's lab) and commercialized by France's Pixium Vision; Science bought the IP and assets out of Pixium's liquidation via the Paris Commercial Court in April 2024 for €4M ($4.7M). So the most valuable clinical asset on the books was acquired for the price of a small house — the upstream "supplier" was a bankrupt estate.
• Capital (the true input): $490M cumulative venture funding — see Lens 5.

The chain is unusually short and owned for a deep-tech medical company: Stanford-origin IP + a captive MEMS foundry + in-house biology. The single-source dependency is itself — concentration risk lives inside the company (one fab, one founder), not in an external vendor.

Lens 3 · Competitive Advantages (moats)

• A shipping product with NEJM-grade efficacy (the only real moat today). PRIMA's PRIMAvera pivotal trial put 25.5 ETDRS letters of mean visual-acuity gain (>5 lines) into the New England Journal of Medicine with 80%+ of patients reading letters/numbers/words. In a field where every other BCI company is pre-revenue and pre-pivotal, "we have peer-reviewed restoration of fluent reading in blind patients" is a category-of-one claim. [durable — clinical data + regulatory head-start]
• Vertical integration / captive MEMS foundry. Owning the fab (ex-MEMSCAP) means process know-how, yield control and IP secrecy that a fab-less rival must rent — a real switching/scale moat for ultra-dense implantable arrays. [durable — process + IP]
• Architectural bet (biohybrid) as a future moat. If living-neuron interfaces work, they sidestep the foreign-body/scarring ceiling that limits electrode longevity — potentially leapfrogging Neuralink's entire approach. But this is an option, not a moat — it is in mice. [speculative]
• Founder pedigree + talent magnet. Hodak (Neuralink co-founder) plus a roster of Neuralink/academic alumni, and named clinical advisers (Yale neurosurgery chair Dr. Murat Günel for the first human biohybrid placement; Sydney Eye Hospital's Dr. Simunovic for RP/Stargardt). [moderate]
• Bargaining power: weak today — pre-commercial, dependent on tier-1 VC capital and on payer reimbursement decisions it does not control. Power accrues only once PRIMA is approved, reimbursed and the surgeon network is trained. Patent estate: ~79 patents filed — the legal moat is being built but is unproven in dispute.

Honest read: the moat is the product's clinical lead and the captive fab, not the moonshot. Bulls who price the biohybrid as a moat today are pricing an option as an asset.

Lens 4 · Segments

n/a — private, not disclosed. There is no segment-level revenue, EBITDA or geographic split because the company is pre-revenue — PRIMA has not launched commercially and the empty segments.csv reflects reality, not a gap in research. The meaningful "segmentation" is by program maturity, which substitutes for a P&L breakdown:

Trend: the center of gravity is shifting toward commercialization — the Series C narrative and the $100M-revenue goal both point at PRIMA as the 2026–27 story, with the moonshots funded off its back.

---

Phase B — Measure performance (+private overlay: SEC-grounded lenses swapped)

Lens 5 · Funding & valuation trajectory (swaps "Earnings Result")

Round history (all ``, unaudited per public sources):

• Series C investors (all returning): Lightspeed Venture Partners, Khosla Ventures, Y Combinator, Quiet Capital, and IQT — the latter being In-Q-Tel, the CIA/intelligence-community strategic fund. The insider-only round (no new lead) is a double-edged signal — see Lens 13.
• Use of proceeds: commercialization of PRIMA, advancing Science Eye + Biohybrid into the clinic, and expanding "research, manufacturing and operational infrastructure" (Biohybrid, Vessel, MEMS Foundry, "BCI Ecosystem").
• Revenue: $0 commercial today; the explicit goal is $100M — a goal, not a guidance figure; treat as aspiration.
• Burn / runway: not disclosed. `` With ~150 staff, a captive NC fab, multi-arm clinical programs and gene-therapy CDMO spend, a plausible burn is ~$80–130M/yr; against the $230M fresh raise that implies ~2 years of runway to an EU launch and first revenue — adequate but not comfortable if PRIMA's reimbursement is slow. Label clearly: this burn band is my derivation, not a sourced number.

Lens 6 · Founder & narrative sentiment trend (calls → founder interviews)

No earnings calls exist; the substitute is Hodak's public narrative across interviews/essays (Futurism, CoreMemory, Arena, company blog):

• 2021–22: "secretive neuroscience startup," consciousness/long-horizon framing — pure moonshot tone.
• 2024: the Pixium acquisition reframes the company around a near-term, regulatable product; biohybrid presented as the long game.
• 2025–26: decisively commercial — "first time restoration of fluent reading has ever been definitively shown in blind patients" (Hodak), the $100M-revenue ambition, "world's most advanced vision restoration technology". The recurring new phrases are commercialization, reimbursement, manufacturing scale; the thing he's stopped saying is the early consciousness-moonshot language. Tone shift: research-lab → medical-device company. That maturation is the single most bullish soft signal in the file.

Lens 7 · Cap table & secondary marks (swaps "Comps")

Syndicate quality (the IPO-proximity tell). Science's backers are top-tier venture but conspicuously not crossover/IPO-stage: Khosla, Lightspeed, Y Combinator, Quiet Capital, IQT. There is no Fidelity / T. Rowe / Coatue / mutual-fund entry — the marker that usually signals IPO proximity is absent. IQT's presence flags strategic/intelligence-community interest (neural interfaces are dual-use), not financial-sponsor IPO prep. No secondary marks or mutual-fund markups are publicly disclosed — n/a — private, not disclosed.

Where it sits vs the BCI field (peer valuations, since there are no public multiples — every name here is private/pre-revenue, so EV/Sales, P/E, ROE are all n/a):

Read: at $1.5B, Science is the #2 BCI valuation behind Neuralink ($9B) but the #1 on clinical/regulatory progress toward an approvable product (PRIMA is past pivotal; Neuralink/Synchron/Paradromics are still in early feasibility). The valuation gap to Neuralink is a brand/scope gap (motor-cortex/consumer BCI TAM dreams), not a progress gap. Total disclosed BCI-sector funding crossed $1.6B in 2025–26. Multiples: n/a (all peers private/pre-revenue).

Lens 8 · Funding & product catalysts (price-moves → milestone events)

There is no stock, so the catalysts are the milestones that re-rated the private mark or will (all ``):

• Apr 2024 — buys PRIMA out of Pixium liquidation for ~€4M → instantly acquires a clinic-stage, FDA-Breakthrough (2023) vision asset. The defining value-creation event.
• Oct 2025 — NEJM publication of PRIMAvera pivotal results (25.5 ETDRS letters) → de-risks efficacy; the scientific credibility unlock.
• Nov 2025 — Hodak's "$100M revenue" / commercialization framing (Axios) → narrative pivot to a business.
• Dec 2025 — Vessel announced → optionality / TAM-expansion signal.
• Mar 2026 — $230M Series C at $1.5B → the financing that funds the EU launch.
• Apr 2026 — first human biohybrid sensor placement prep with Yale's Günel → the moonshot starts its clinical clock.

Pattern: clinical/regulatory de-risking drives the mark, not hype. This name re-rates on data readouts and approvals — the most fundamentally-driven profile in the BCI cohort.

Lens 8b · Traction & unit economics (+private add-on)

• Commercial traction: zero revenue; PRIMAvera enrolled 38 patients (geographic-atrophy), 47 across EU+US programs cumulatively. New trials launching for Stargardt disease and retinitis pigmentosa at Sydney Eye Hospital.
• Unit economics (``, unaudited): PRIMA is a one-time implant + glasses/software. Comparable advanced retinal/neuro-implants historically priced $100K–$150K+ per device/procedure (e.g. the legacy Argus II retinal prosthesis listed ~$150K). At a ~700K US GA-prevalence pool, even a low-single-digit-percent penetration is a multi-hundred-million-dollar revenue line — which is what makes Hodak's $100M target arithmetically reachable. The pricing/penetration figures are my estimate from market comps, not company-sourced; treat as illustrative.
• Gross-margin signal: captive MEMS fab + a high-ASP implant should support medtech-class gross margins (60–80%) at scale `` — but yield on ultra-dense photovoltaic arrays is the swing factor and is undisclosed.

---

Phase C — Judge people & books

Lens 9 · Management

• Max Hodak — Founder & CEO. Track record: co-founded Neuralink (2016) and served as president before departing May 2021; earlier founded Transcriptic (lab-automation). He is a genuine 0→1 neurotech operator. Skin in the game: founder, presumably large equity holder (undisclosed). Capital allocation — and this is his standout signal: the $4.7M Pixium buy is one of the best venture-asset trades in recent medtech — acquiring a clinic-stage, FDA-Breakthrough vision platform out of bankruptcy and turning it into an NEJM result and a $1.5B valuation in under two years. That is elite capital allocation. Red flags: the circumstances of his Neuralink exit were never disclosed — a soft governance question; and a multi-front strategy (vision + biohybrid + perfusion + foundry) on $490M risks focus dilution (see Lens 13). Archetype: founder-visionary — appropriate for this frontier stage, but the field-defining test is whether he can now run a disciplined commercial org (manufacturing, reimbursement, sales) rather than a research lab.
• Key technical leadership: Alan Mardinly (biohybrid program lead, quoted on the living-neuron architecture); clinical advisers Dr. Murat Günel (Yale, neurosurgery chair) and Dr. Matthew Simunovic (Sydney Eye Hospital).
• Insider ownership / comp: n/a — private, not disclosed.

Lens 10 · Forensic Red Flags

Accounting: n/a — no audited financials exist (private, no SEC filings; financials.csv empty). There is nothing to forensically analyze on the income statement / balance sheet / cash flow — and that itself is the headline risk for an investor: zero financial transparency, unaudited, illiquid. Specific watch-items for when disclosure eventually arrives: (1) R&D capitalization vs expensing (a multi-program deep-tech co. can flatter losses by capitalizing); (2) the carrying value/impairment of the Pixium-acquired IP (bought for ~$4.7M, now central to a $1.5B valuation — a future audit must justify intangibles); (3) fab depreciation on the $65M NC foundry investment.

Regulatory findings (required sub-section). Per regulatory/regulatory-findings.md (Step 0, fetched 2026-06-30): Science Corporation has no CIK — it is private and not required to file with the SEC, so no SEC Litigation Releases or AAERs are possible (total_sec_findings: 0).

• Non-SEC enforcement web search ("Science Corporation" (FTC OR DOJ OR FDA OR CFPB OR settlement OR fine OR penalty) enforcement): no material adverse hits. No lawsuits, FDA warning letters, patent-dispute litigation, or consent decrees surfaced. To the contrary, PRIMA holds FDA Breakthrough Device designation (2023) and the company is progressing CE-mark/FDA applications normally.
• There is no 10-K Item 3 to quote (private).
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (no-CIK ⇒ N/A), web search, and the absence of any public enforcement/litigation record as of 2026-06-30. Clinical-safety caveat (not a regulatory finding): PRIMAvera reported serious surgery-related adverse events — retinal detachment, choroidal neovascularization, vitreous hemorrhage, transient IOP spikes. These are device/procedure-risk disclosures, material to the regulatory path, and belong in the bear case (Lens 12/13).

---

Phase D — Project & stress-test

Lens 11 · IPO-readiness & path-to-tradeable (swaps "Forward Projection")

This is the be-early payoff lens. Pulling the structural fields (none exist yet for science-corp in private-watch.json — write-back proposed below):

• Stage: growth → late (clinical-stage with a pivotal-complete lead product, $490M raised, $1.5B mark).
• IPO readiness: 2 / 5 (1=seed … 5=S-1 filed). Rationale: progress is real but the investor signature is wrong for IPO proximity — an insider-only Series C with no crossover/mutual-fund participation (no Fidelity/T. Rowe/Coatue) says the syndicate is funding the private commercialization phase, not staging a public exit. Compare Neuralink (readiness ~4, QIA/Thrive/Sequoia/ARK breadth).
• Milestones that unlock an S-1: (1) EU/Germany PRIMA approval + first commercial implants (~mid-2026); (2) FDA approval in the US (timeline undisclosed, post-EU); (3) demonstrated revenue ramp toward the $100M goal + a reimbursement framework; (4) a crossover round as the tell. Estimated tradeable window: not before 2028, contingent on US approval and a revenue base ``.
• No EPS projection is logged — pre-revenue private; an EPS forecast would be fabrication. Per SKILL +private + --watchlist rules, no forecast.ts create is run. The scoreable binary, if Connor wants one later, is "PRIMA receives EU CE-mark / first commercial approval by 2026-12-31" — high probability given filings are in ``.
• private-watch.json write-back (proposed; not executed in this wave per the wave-boundary rule):

"science-corp": {
"beat": "bci", "stage": "late", "ipo_readiness": 2,
"lead_investors": "Khosla, Lightspeed, Y Combinator, Quiet Capital, IQT",
"catalyst": "PRIMA (vision-restoration retinal implant) — pivotal NEJM data (25.5 ETDRS letters); CE-mark/FDA filed, EU launch ~mid-2026; biohybrid cortical interface in mice. Insider-only Series C ($1.5B) — no crossover yet, so pre-IPO is >2yr out.",
"dossier": "../menfem-research/companies/science-corp/deep-dive-2026-06-30.md"
}

Lens 12 · Bull vs Bear

Bull case. Science is the only BCI company with a real, peer-reviewed, near-approvable product — PRIMA restored fluent reading to blind dry-AMD patients (25.5 ETDRS letters, NEJM), a category-defining clinical result in a 700K-US-patient / $20B+ market with no curative competitor (the approved GA drugs only slow decline). It owns its MEMS fab, bought its lead asset for a rounding-error $4.7M, and is run by a founder with elite asset-allocation instincts. The $1.5B mark is ~6x below Neuralink's $9B despite being further along the regulatory path. The optionality is enormous: if the biohybrid interface works, Science leapfrogs the entire electrode-based field. You are buying a funded medical-device cash engine wrapped around a free moonshot call option.

Bear case (2–3 permanent-impairment risks). (1) PRIMA's commercial reality may disappoint the clinical drama — a $100K+ surgical implant for elderly GA patients faces reimbursement friction, surgeon-training bottlenecks, and serious surgical AEs (retinal detachment, CNV, vitreous hemorrhage) that could throttle adoption far below the headline TAM. (2) Focus dilution — vision + biohybrid + Vessel + foundry on $490M is a lot of fronts; a private with no public cash discipline can burn its runway chasing four moonshots and need a down-round. (3) The biohybrid is in mice — a decade and many biology miracles from human cognitive restoration; pricing it as anything but a lottery ticket is a mistake. Pre-mortem (18 months out, thesis broke): EU approval slipped or came with a narrow label; first commercial implants revealed real-world durability/AE problems; revenue stayed near zero; the next round was a flat-or-down insider bridge; and the biohybrid human placement produced no usable signal. Multiples too high? Unknowable (private, no comps) — but $1.5B on $0 revenue is a clinical-milestone price, not a fundamentals price; it re-rates only on approval + revenue. Contrarian view the market refuses to see: the vision business — not the cortical moonshot everyone fixates on — is the actual company, and it should arguably be valued as a high-growth ophthalmic medical-device play (where comps are far below $1.5B on $0 sales), with the biohybrid as upside.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case. Revenue concentration is total and binary — 100% of the near-term thesis rides on one product (PRIMA) getting approved AND reimbursed AND adopted in elderly surgical patients. If German payers balk at a six-figure implant, or surgical-complication rates scare the surgeon community, the commercial engine stalls and the $1.5B has nothing under it. The moat is thinner than bulls think: PRIMA's core IP came from Stanford via a bankrupt French company — Science is the third owner of a technology that Pixium itself could not commercialize into solvency (it liquidated). Why does Science succeed where the originator went bankrupt? The answer ("better capital, manufacturing, timing") is plausible but unproven. The most dangerous competitor bulls underestimate: not Neuralink — it's pharma. Apellis and Astellas/Iveric already sell injectable, no-surgery GA drugs reimbursed today; if next-gen complement or gene therapies restore (not just slow) vision without an eye implant, PRIMA's surgical proposition is structurally disadvantaged. Capital-allocation worry: the insider-only Series C with no new lead and no crossover can be read two ways — loyalty, or "no outside lead would price it higher," a classic late-private warning. Assumptions that must hold for $1.5B: EU approval on time, US approval following, six-figure pricing accepted by payers, manageable AE profile in commercial use, and continued cheap capital. If growth disappoints 20–30%: for a pre-revenue name there is no "growth" to miss — the asymmetric risk is a delayed/narrow approval, which would trigger a down-round since the mark is pure milestone-credit. Single scenario that permanently impairs: a serious-safety signal in commercial PRIMA use (a cluster of vision-threatening complications) that forces a label restriction or withdrawal — plausibility moderate given the trial already logged retinal detachments and CNV.

Lens 14 · Management Questions (ordered by information value)

1. PRIMA's reimbursement path — what is the expected per-implant price, and which German/EU payers have committed to coverage at launch? (The whole near-term thesis hinges on this.)
2. What were the rates and severity of serious adverse events (retinal detachment, CNV, vitreous hemorrhage) in PRIMAvera, and how do they trend with surgeon experience?
3. What is the company's annual burn and current runway, and at what revenue/milestone does the $230M get you to the next financing?
4. Why did you raise an insider-only Series C with no new lead or crossover investor — and when do you expect a crossover round?
5. What US FDA pathway and timeline does PRIMA follow post-EU, and what did the agency signal in your pre-submission meetings?
6. Pixium could not commercialize PRIMA into solvency — concretely, what do you do differently (cost, manufacturing, indication) that makes it viable?
7. How do you defend against pharma (Apellis/Astellas) and future non-surgical GA gene therapies that could restore vision without an implant?
8. With four programs (PRIMA, Science Eye, Biohybrid, Vessel) on $490M, how do you prioritize capital, and what gets cut if PRIMA's launch is slow?
9. The biohybrid is in mice — what are the specific, dated milestones (channel count, signal fidelity, primate data) before a useful human cortical interface, and what would make you kill it?
10. What is the manufacturing yield on the photovoltaic implant array at the Science Foundry, and what gross margin does that imply at commercial volume?
11. What durable IP / patent protection covers PRIMA and the biohybrid scaffold, and what is your freedom-to-operate position vs Stanford-origin and competing patents?
12. How do you train and scale the surgeon network required to implant PRIMA at volume across the EU and US?
13. What is the realistic addressable patient count (vs the 700K GA prevalence) given surgical eligibility, and your year-3 implant-volume target?
14. What are the circumstances of your departure from Neuralink, and what did it teach you about running this company?
15. What does success in five years look like in revenue and approved indications — and at what point would you consider an IPO?

---