30-May-2026 Pitch Deck Review 2

posted: 30-May-2026 & updated: 30-May-2026

Erudio Bio Seed Pitch Deck — v2 Review

Deck: Erudio Investor Pitch 2026 05 30.pdf (24 slides, 15.1 MB, dated 2026-05-30)
Prior review: output/pitch-review.md (v1 dated 2026-05-25, 12 slides)
Date of this review: 2026-05-30\

What changed since v1: the user expanded the deck from 12 → 24 slides, applying many but not all of the v1 recommendations. New material concentrated in Slides 2–7, 12, 13, 14, 16, 21. Carry-over slides include the Biodefense bullets (Slide 22) and the hero validation chart (Slide 18). The user noted up-front that ask / use-of-funds / milestones are not yet added.

1. Headline

Reviewer	v1 grade	v2 grade	Movement
Narrative	C-	B-	Real structural arc now exists (problem → vision → tech → biz model → competition → applications → team). Slide 1 still wasted; three-market focus problem unresolved.
Numbers	D+	C+	Gates grant + named JDA + $60B TAM + named competitors are real wins. New “1000x”, “ADI IP”, and “$60B TAM” all need sourcing. Slide 18 hero chart unchanged.
Devil’s Advocate	REVISIT IN 6 MONTHS	REVISIT IN ~3 MONTHS	The deck is now memo-able for a soft-pass. Still not investable today (no ask, no powered benchmark, no commercial hire).

2. Where the revisions clearly landed (big wins)

NEW Slide 2 Executive Summary carries the hook the previous Slide 1 was missing. Gates Foundation $1M grant + named JDA negotiations with SNUH/Keimyung + $60B TAM + bioTCAD framing all show up. This is the most important single change in the deck.
NEW Slide 3 (“Why Do We Make Life-Altering Decisions with Incomplete Data?”) — real problem statement with a cited source (Bains/DDW), framing the “cross-reactivity / limited testing / doctor-deduction-bottleneck” pain in plain English.
NEW Slide 16 (“Drug Development is Very Inefficient — Fails 90% and Costs $2.6B”) — quantified pain slide for bioTCAD specifically. Names the Hit → Lead → Preclinical → IND → Clinical funnel. The “in-vitro / in-silico / in-vivo” framing is sharp.
NEW Slide 13 Business Model — razor/blade + SaaS, 70% margin claim, named revenue components (instrument, chip + flowcell, reagent kit, software). Tells the investor how Erudio gets paid, which v1 never did.
NEW Slide 14 Competitive Landscape — names real competitors (Luminex, Olink, Lino Biotech, ProteinSimple, Depixus, LUMICKS). v1 had none. (Caveats in §4.)
Slide 11 retool — the cytokine-immunoassay slide now has IL-6 / IL-8 / TNF-α actually shown, a 3-panel “Challenge → Filtering → Corrected Multiplex Readout” story, and the punchline “Cross-reactivity removed. Multiplex performance restored.” This is the cleanest narrative arc on the deck. Numbers reviewer specifically called this slide more defensible than the bioTCAD hero chart — you’ve now earned that.
Slide 21 Drug Development engine diagram — replaced the v1 box diagram with a real input/processing/output engine schematic. Named outputs: binding prediction, functional simulation, candidate prioritization, experimental guidance.
Slide 22 Biodefense visual upgrade — ruggedized hardware photo, “DOZENS, HUNDREDS, THOUSANDS!” scalability bar, FIELD HEALTH AGENCY-style insignia, “ONE PLATFORM, ONE SAMPLE” lifecycle bar (Pre-deployment / Deployment / Post-deployment / Core Technology). Visually the strongest slide in the deck. (Bullet text still needs updating — §3.)
Slide 23 Clinical Diagnostic visual upgrade — hospital workflow, SNUH and Keimyung logos in color, “PREVENTATIVE SCREENING / DIAGNOSTIC WORKUP / LONGITUDINAL MONITORING” core-tech bar.

3. Carried-over issues that did NOT get fixed (from v1 list)

Fix from v1 review	Status in v2
#1 Add Ask slide / use of funds / milestones	Missing (user-acknowledged)
#2 Pick one wedge	Not done. Tech Stack (Slide 6) and Applications section (Slides 20–23) still pitch three markets.
#3 Rewrite Slide 1	Not done. Slide 1 is still the generic logo + tagline. The hook moved to Slide 2; Slide 1 itself is still a wasted slide.
#5 Name Schrödinger et al. in competitive view	Partial. Slide 14 names assay competitors. No competitive view for bioTCAD (Schrödinger FEP+, Iambic, Isomorphic, Recursion, Insilico).
#6 Replace Slide 7/18 hero chart with powered benchmark	Not done. Slide 18 is unchanged. Same n=8, same unnamed comparator, same anchor-in-train-set issue.
#8 Anchor IP claim with USPTO numbers	Not done. Slide 9 still “21 issued patents” with no table. Made worse by new ADI claim — see §4.1.
#9 Label Synopsys/Samsung/SK Hynix logos	Not done. Slide 19 still has them unlabeled at the bottom. Same logos now also reappear on Slide 24 (Team) without labels.
#11 Soften absolute language	Not done. “Error-free multiplexing” still on Slide 9. “Requirement for every launch” still on Slide 17. “20% of world’s economy” appears on two slides (5 and 23) — propagation, not retraction.
#12 Commercial hire on team	Not done. Team unchanged.
#13 Disclose Carterra conflict	Not done.
Carry-over	Slide 22 Biodefense bullets still the old “GWOT to Great Power Competition / Need to defend / Flexible, efficient solution from CBRND to Readiness” — exactly the text Option B (in our last exchange) rewrote. Just hasn’t been applied yet.
Carry-over typo	Slide 23 still says “single sayer networks” — should be “single-payer networks”.

4. New issues introduced by the v2 additions

4.1 The “Analog Devices” claim on Slide 2 creates IP confusion

“Part of IP Licensed from ADI” “Erudio Bio and Analog can be a key player in a fast-growing high margin business at the cross between biotechnology and AI.”

Conflicts with the “Stanford spin-out” framing still on Slide 9. Is the core IP Stanford-licensed, ADI-licensed, or Erudio-assigned?
“Erudio Bio and Analog can be a key player” — grammar issue, and “Analog” alone is not a brand.
No date, no scope, no exclusivity. “Part of IP” is unusually vague.

See §7.3 for a rewrite template.

4.2 “1000x more AI-enhanced biodata” on Slide 2 is exactly the kind of number that gets killed

1000x what? Vs. which baseline? Per sample? Per dollar? Per hour? Same problem as v1’s “MASSIVELY multiplexed” — unsourced order-of-magnitude on the front. Either source it (“1000x more biodata per microliter of sample vs. Luminex xMAP”) or change to a sourced comparison.

4.3 Slide 14 Competitive Landscape — right idea, wrong execution

The 2×2 “Data amount” × “Data quality” with Erudio alone in the upper-right is the most clichéd competitive-slide format. Every investor who has seen >50 decks is trained to be skeptical when the pitching company sits alone in the magic quadrant.
Competitor set is incomplete. Listed: Luminex, Olink, Lino Biotech, ProteinSimple, Depixus, LUMICKS — all multiplex assay / single-molecule biophysics. This covers VSA but not bioTCAD. Need a second slide or split landscape for Schrödinger (FEP+), Iambic, Isomorphic Labs, Recursion, Insilico, Chai.
Carterra is missing from the competitor list — and Tim Germann (Carterra CCO) is on the advisory board. That conflict is now even more conspicuous.

4.4 Executive Summary (Slide 2) is overloaded

Six headline claims on one slide. Even sophisticated readers can absorb at most three. Ranking:

Keep: Gates Foundation $1M grant (third-party validation, real money).
Keep: bioTCAD framing line (“measurement-calibrated, physics-based drug discovery platform”).
Keep: Hospital JDA in advanced negotiation (real, specific, signed-deal-adjacent).
Demote to IP slide: the ADI line — and rewrite (§4.1).
Demote or remove: “1000x” claim (until sourced — §4.2).
Demote to market slide: $60B TAM (needs methodology — full drug-discovery R&D spend? FEP/MD software market? Addressable-to-bioTCAD SAM? Big difference).

See §7.2 for a rewritten Exec Summary.

4.5 Slide 5 Global Urgency now states “20% of world’s economy” as a giant 20% callout

You upgraded the visual weight of the figure that the numbers reviewer flagged as overstating reality by ~2× (WHO/OECD: global health spend ~10% of world GDP; ~17% is US). When the number is wrong, making it bigger makes it worse. Defensible replacement: “Global health spending reached ~$9 trillion in 2023 (≈10% of world GDP, growing ~4% real per year) — WHO GHED.”

4.6 Slide 17 bioTCAD intro now has a side-by-side “semiconductor TCAD ↔ pharmaceutical bioTCAD”

Strong visual addition. But the slide still keeps “bioTCAD will be as widely used as semiconductor EDA: a requirement for every launch” as the punchline. The visual makes the analogy compelling; the bold punchline makes it risky. Soften to “bioTCAD aims to become a standard tool for medicinal-chemistry teams developing novel modalities.”

4.7 Slide 6 Tech Stack — improved but still pitches three markets

The revised Tech Stack is genuinely better (named applications, cleaner layering). But by naming Doctor-in-a-box / bioTCAD / Clinical Diagnostics / Biodefense / Drug Development explicitly, it locks in the three-market sprawl right at the top. If you accept the v1 recommendation to focus the VC deck on drug development, this slide needs a single-wedge version.

4.8 Section dividers (Slides 8, 15, 20)

Three nearly-blank section dividers in a 24-slide deck = ~12% of slides are navigational. Fold into the section’s first content slide; saves 3 slides.

5. Prioritized edits for the next pass

Stack-ranked by investor impact, accounting for ask/use-of-funds/milestones being already on the queue.

Fix the ADI / Stanford IP story (§4.1, copy in §7.3). Until clear, the deck reads as “core IP unowned.”
Rewrite Slide 1. Borrow from the strongest Exec Summary line + Gates grant. (Copy in §7.1.)
Source or replace “20%”, “1000x”, and “$60B” (§4.2, §4.5). Three citations. Half a day of work.
Apply the Biodefense Option B bullet rewrite to Slide 22. (Reproduced in §7.4 for convenience.)
Fix Slide 14 Competitive Landscape (§4.3) — drop the 2×2 cliché; add bioTCAD competitors; include or differentiate Carterra.
Replace Slide 18 hero chart with a powered benchmark, even if interim. If you can’t get n=30 yet, at least re-caption honestly.
Soften absolutes — “error-free” → “error-suppressed” (Slide 9); “a requirement for every launch” → “a standard tool for novel-modality medchem” (Slide 17).
Label or remove the Synopsys/Samsung/SK Hynix logos on Slides 19 and 24.
Tighten the Executive Summary to three headline claims (§4.4, copy in §7.2).
Fix the “single sayer networks” typo on Slide 23 → “single-payer networks.”
Disclose the Carterra relationship — either on the competitive slide as differentiation, or as a footnote on Tim Germann’s advisory listing.
(Acknowledged, coming): Ask slide + use of funds + 12/18/24-month milestones.

6. Two structural questions worth answering before the next pass

Is this the VC seed deck, the strategic/government capability brief, or both? v1 recommended a two-deck strategy. v2 is still trying to be both. The single biggest tightening lever is to commit to one.
What is the Erudio / ADI / Stanford relationship in one sentence? This is now ambiguous in a way that v1 was not. The ADI line is potentially a huge asset (strategic partner, distribution channel, manufacturing access) — but only if it’s stated unambiguously.

7. Copy-paste-ready text for the next pass

The text below is ready to type directly into PowerPoint. Bullet markers will be added by the slide layout; copy text only.

7.1 Slide 1 — Hook (replaces generic logo + tagline)

Pick one of these three. Title remains “Erudio Bio” with the logo; the second line replaces “Data Driven AI for Effective Medicine.”

Option A — concrete, three proof points (recommended)

Measurement-calibrated AI for drug discovery and clinical diagnostics. Backed by the Bill & Melinda Gates Foundation. Clinical JDA in advanced negotiation with Seoul National University Hospital and Keimyung University Dongsan Hospital.

Option B — tighter, two proof points

Measurement-calibrated AI for drug discovery and clinical diagnostics. Gates-Foundation-backed. Clinical JDAs in advanced negotiation in South Korea.

Option C — bioTCAD-first (if you commit to drug-development-as-wedge for the VC deck)

bioTCAD — measurement-calibrated, physics-grounded AI for medicinal chemistry in novel chemical space. Backed by a $1M Bill & Melinda Gates Foundation grant.

7.2 Slide 2 — Tightened Executive Summary (3 claims, not 6)

Replace the current 6-bullet Executive Summary with this. Four labeled sections; only one claim per section.

WHAT WE DO Erudio Bio combines a chip-based biophysics platform (VSA) with a hybrid physics-grounded AI engine (bioTCAD) to generate high-fidelity biological data at scale and to predict molecular behavior in novel chemical space — the regime where AI-only models break down.

THIRD-PARTY VALIDATION $1M Bill & Melinda Gates Foundation grant validating bioTCAD for early-stage drug discovery.

COMMERCIAL TRACTION Joint Development Agreement in advanced negotiation with Seoul National University Hospital and Keimyung University Dongsan Hospital — preventative screening as the first clinical wedge.

WHY NOW Drug discovery still fails 90% of candidates and costs ~$2.6B per approval; AI-alone approaches have hit a credibility wall outside familiar chemical space — exactly what bioTCAD’s measurement-anchored hybrid approach is designed to close.

Items demoted off this slide:

“Part of IP Licensed from ADI” → moves to a dedicated IP slide (§7.3 copy).
“1000x more AI-enhanced biodata” → removed until sourced.
“$60B drug discovery TAM” → moves to the competitive/market slide with WHO/Evaluate/IQVIA-style methodology.

7.3 IP-clarification one-liner (for Slide 9, or a new dedicated IP slide)

The right text depends on the actual relationship between Erudio, Stanford, and Analog Devices. Pick the variant that matches reality and fill in the brackets. If none matches, write a new sentence with the same structure (who owns, who licenses, what scope, what exclusivity).

Variant 1 — Stanford-licensed core + ADI as additional licensor

Core IP: [N] issued patents exclusively licensed from Stanford University covering VSA chip architecture and force-spectroscopy methods (10-year R&D heritage). [M] additional patents licensed from Analog Devices covering [specific tech area, e.g., readout electronics / sensor-array addressing] for bio-sensing field-of-use.

Variant 2 — Stanford and ADI co-licensors of a single tech bundle

Core IP: chip-architecture, actuation, and readout patents licensed from Stanford University and Analog Devices, with Erudio Bio as the exclusive bio-applications licensee for [scope, e.g., diagnostics + drug discovery].

Variant 3 — Erudio-assigned, with prior ADI know-how / inventor history

Core IP: [N] issued patents assigned to Erudio Bio, building on inventor experience at Analog Devices and 10 years of development at Stanford. Filed in [US / PCT / KR / EU].

Whichever you pick, three things must appear in one sentence somewhere on the deck: assignee / licensee, scope, exclusivity. Without those, “Part of IP Licensed from ADI” reads as “we don’t own our core IP.”

Also drop the second ADI sentence (“Erudio Bio and Analog can be a key player…”) entirely from the Exec Summary — it’s not a fact, it’s a hope. If ADI is a strategic partner with named commitments, that goes on a Partnerships slide with terms.

7.4 Slide 22 — Biodefense bullets (Option B from prior exchange, restated)

To replace the current “GWOT to Great Power Competition / Need to defend against near-peer adversaries / Flexible, efficient solution needed from CBRND to Readiness.”

Bullet 1

Peer adversaries now field engineered, signature-unknown bio/chem agents. The threat library is no longer the threat.

Bullet 2

Legacy assays answer “is it X?” Erudio answers “what is it, even if it’s new?”

Bullet 3

One chip, sample-to-answer, ruggedized for the forward edge — covering the CBRND-to-readiness continuum on a single platform.

7.5 Slide 5 — sourced replacement for “20% of world’s economy”

Replace the giant “20%” callout with:

Global health spending reached ~$9 trillion in 2023 — ≈10% of world GDP, growing ~4% in real terms per year (WHO Global Health Expenditure Database).

Per-capita spending growth in advanced economies is outpacing GDP growth, driven by aging populations.

National health systems treat cost containment as an existential issue.

Efficient preventative screening is widely seen as the structural solution.

Same urgency; sourced figure. The “outpacing EU GDP by 3×” line can stay if you can cite a per-capita-growth-rate source for it.

7.6 Two-word fixes (drop-in)

Slide 9, body text: “Error-free multiplexing” → “Error-suppressed multiplexing”
Slide 17, punchline: “a requirement for every launch” → “a standard tool for novel-modality medicinal chemistry”
Slide 23, body text: “single sayer networks” → “single-payer networks”

8. What I did not draft (waiting on your input)

Ask slide / use of funds / 12-18-24mo milestones — you said this is coming. Happy to draft a structure when you’re ready.
Slide 14 Competitive Landscape rewrite — depends on whether you want a feature-comparison table or a quadrant chart on different (real, measurable) axes. Tell me which and I’ll draft it.
Slide 18 hero benchmark rewrite — depends on whether you have new validation data or just want to re-caption the existing chart honestly. Tell me which.
IP slide full draft — depends on the actual Erudio/Stanford/ADI relationship. Tell me the real structure and I’ll write the slide.

Erudio Bio, Inc.