Cognixion

Phase A — Understand the business

Company Overview

Cognixion is a Santa Barbara, CA assistive-technology company building a non-invasive brain-computer interface (BCI) fused with augmented reality (AR) and a generative-AI communication engine, aimed at people who have lost the ability to speak — primarily late-stage ALS and locked-in patients. Founded 2014 by Andreas Forsland; it is the second act of his earlier startup Smartstones (founded 2013), which is now "permanently closed" per Crunchbase — i.e. Cognixion is a pivot/relaunch of the same founder's "think-to-speak" mission, not a clean-sheet company.

What it actually makes / sells — two distinct lines:

1. Cognixion ONE / Cognixion ONE Axon — the flagship medical device: a wireless AR headset that reads EEG over the occipital lobe, detects steady-state visual evoked potentials (SSVEP), overlays selectable letter/word/phrase targets in AR, and runs a per-patient LLM "conversational co-pilot" trained on the user's own past speech/writing to reach "near-conversational" communication. This is the FDA Breakthrough Device candidate (see Lens 10) — not yet FDA-cleared; it is being validated in an ALS trial.
2. Cognixion ONE Axon-R — a research-grade version of the same platform, sold to universities/hospitals/labs as a multi-modal (EEG + AR + AI) neural-interface research hub, up to 16 channels with a built-in SSVEP classifier. Explicitly "research-grade only (not FDA-cleared)." Starting price $25,000.

Business model. B2B + direct: subscription/partnership agreements with healthcare institutions for the assistive line, plus hardware sales of Axon-R into research. In practice the only near-term, clearable revenue is the $25K Axon-R research tool — the medical device is pre-clearance and therefore pre-meaningful-reimbursement.

Customers / channel. Research institutions reached via two distributors (see Lens 2). The eventual assistive-device customer is the ALS/locked-in patient, funded — if cleared — through medical/durable-medical-equipment channels that today pay for eye-tracking speech devices (Tobii Dynavox).

Verdict on the model: a genuine, sympathetic clinical problem with a credible non-invasive angle, but a narrow wedge (advanced ALS where even eye-tracking fails) bridged by a small research-tools business to keep the lights on. The TAM story and the revenue story are not the same company yet.

Supply Chain

Map: EEG/AR/compute components → Cognixion (designs + assembles the headset + software) → distributor → research lab / (future) patient.

Named stakeholders along the chain:

• Upstream (inputs): Not publicly disclosed at the component level [n/a — private, not disclosed]. The headset integrates dry/standard EEG electrodes, an AR optical display, onboard compute, and a battery; Cognixion positions itself as the systems integrator + software/AI owner, which is where its value concentrates (the SSVEP classifier and the per-patient LLM co-pilot are proprietary).
• AI layer: a "conversational co-pilot" generative-AI model fine-tuned per patient on their own speech/writing samples. The underlying foundation model is not disclosed — a dependency worth flagging (third-party LLM API vs. in-house).
• Distribution (the important part):
• BIOPAC Systems — distributes and supports Axon-R for the research market; markets it as a 16-channel immersive BCI platform.
• Blackrock Neurotech — as of May 2025, a non-exclusive distributor of Axon-R to research institutions via its network; Blackrock separately builds implantable BCIs, so Axon-R is a complementary non-invasive offering in its catalog.
• Platform integration: Cognixion has demoed combining its BCI with Apple Vision Pro — a tell that it wants to ride third-party AR hardware rather than carry the full cost of bespoke optics forever.

Chokepoints / single-source dependencies:

• Distribution is outsourced, not owned. Both research-channel partners are non-exclusive; Cognixion does not control the customer relationship. Good for capital-light reach, bad for margin capture and durability — a partner can drop the line.
• Clinical/regulatory throughput is the real bottleneck, not parts. The binding constraint is FDA clearance + reimbursement, not supply (see Lens 10/11).

Names or it didn't happen — secured: BIOPAC, Blackrock Neurotech, Apple (Vision Pro integration), ALS Association (trial recruitment). Component suppliers remain undisclosed.

Competitive Advantages (moats)

The differentiator — non-invasive + AR + per-patient AI. Against the headline BCI names (Neuralink, Synchron, Precision Neuroscience) the entire pitch is: no surgery, no implant, no neurosurgeon, works today out of a box, and reaches patients after eye-tracking fails. For the specific advanced-ALS/locked-in segment that has lost ocular motility, that is a real, defensible position the implant players can't cheaply contest, and the eye-tracker incumbents (Tobii) physically can't serve.

Candidate moats, graded:

• Regulatory head-start (moderate, real): FDA Breakthrough Device designation (2023) for a non-invasive BCI for assistive communication — Cognixion claims it aims to be the first FDA-cleared non-invasive BCI letting fully locked-in patients communicate. Breakthrough status is a process accelerator and a credibility marker, not a clearance and not an exclusivity grant.
• IP / SSVEP + AR + LLM integration stack (moderate): the combination — AR target presentation at distinct flicker frequencies + occipital SSVEP decode + a patient-specific language model — is a non-trivial systems integration. Patents not verified in this pass [n/a].
• Mission/brand & founder story (weak-but-real for this niche): a deeply sympathetic, founder-led origin (Forsland's mother) gives it standing with the ALS community, clinicians, and impact investors (CABHI, Verizon Forward for Good, Amazon Alexa Fund).
• Switching costs / network effects (absent today): none meaningful pre-scale.

Bargaining power. Weak on both sides. Over suppliers: it's a tiny buyer of commodity-ish AR/EEG parts. Over customers: it sells through other people's distribution (BIOPAC, Blackrock) into a research market and, on the clinical side, into a reimbursement system it doesn't control. The moat is the clinical niche + the regulatory wedge, not pricing power.

Segments

No segment financials exist — private, our figures empty ``. Qualitative split:

Geography: US-centric (Santa Barbara HQ; LA-area trial; US research channel). No international breakout [n/a — private, not disclosed].

The honest read: revenue is immaterial and undisclosed. The "company" today is a clinical program with a small research-hardware side business, not a segmented operating business.

---

Phase B — Measure performance

Funding & valuation trajectory

No earnings exist. The scoreboard is rounds + valuation — and the sources materially disagree. Both are reported; the conflict is itself the finding.

Source A — Tracxn (2026): ~$25M across 8 rounds:

*Several "post-money" figures (e.g. $1.6M post on a ~$9M round; a $6.5M post-money on the Series A) are internally implausible and almost certainly mis-parsed by the aggregator — treat Tracxn's valuation column as unreliable.

Source B — other aggregators (Tracxn/PitchBook-derived, 2026): "$32.22M over 11 rounds," latest a "Convertible Note IV, $4M, Feb 2026," valuation $45.5M as of Feb 2026.

Source C — primary, well-attested: November 2021 ~$12M seed led by Prime Movers Lab, with Amazon (Alexa Fund), Northwell Holdings, Volta Circle. (Note this $12M is larger than any single 2021 round in Tracxn's table — another reason to distrust the round-by-round aggregator detail.)

What is safe to conclude:

• Total raised ≈ $25–32M over ~8–12 years — modest, and increasingly debt/convertible-note financed (Oct 2023, Jan 2026, and a reported Feb 2026 note). Leaning on convertibles late is a classic sign of a priced equity round being hard to get — i.e. flat-to-soft valuation and bridge financing, not a hot up-round.
• Implied valuation is roughly $10M–$46M depending on source — i.e. sub-$50M after 12 years. For context, that is ~1/200th of Neuralink and below even a single early Synchron round (see Lens 7).
• Burn signal: ~29–32 employees on ~$25–32M cumulative over a decade ⇒ low single-digit-million annual burn ``. Survivable, but it explains the convertible-note drip.

Investor base (named, attested): Amazon Alexa Fund, Prime Movers Lab, Northwell Holdings, CABHI, Memorial Hermann Foundation, Verizon Forward for Good, Volta Circle, angel Tom Washing. Read the syndicate honestly: this is a strategic/impact/health-system cap table (a health system, two foundations, a telco CSR fund, an Amazon device-ecosystem fund), not a tier-1 crossover venture syndicate. There is no Fidelity / T. Rowe / Coatue / a16z-growth name — the IPO-proximity tell is absent.

Founder & narrative sentiment

No earnings calls. Proxy = founder interviews + press tone over time:

• Consistent, mission-locked messaging for a decade: "think to speak," "awaken millions," communication as a human right. Forsland is a designer-founder (Apple, IBM, Philips, Citrix design roles) who tells a product-and-empathy story, not a deep-neuroscience or commercial-scale story.
• Tone shift 2023→2025: from "vision/breakthrough" (FDA designation, 2023) toward "now available / commercial" language around Axon-R and the Blackrock/BIOPAC deals (2025), and "conversation-level communication" framing for the trial launch. That is the right direction — evidence-and-distribution over hype — but it's also the pivot of a company that needed a revenue story (research tools) while the clinical story plays out slowly.
• What they've stopped emphasizing: the splashy consumer-AR "world's first BCI+AR wearable" framing of ~2020–2021 has narrowed to a focused ALS/locked-in clinical wedge + a research-tools SKU. Healthy de-scoping, but also a tacit admission the broad consumer dream didn't fund itself.

Cap table & secondary marks / mechanism comps (+private)

There are no public secondary marks or mutual-fund markups for Cognixion [n/a — private, not disclosed]. The instructive comp is the BCI peer set — and the gap is the whole story:

Non-invasive peers (the truer comp set, mostly research/consumer): Emotiv, OpenBCI, Neurable, BrainCo, g.tec, Neuroelectrics, MindMaze; NextMind was acquired by Snap (2023). Non-invasive devices hold ~75% of BCI units but a small slice of value.

The read: Cognixion is one-to-two orders of magnitude smaller than the invasive leaders and sits in the lower-value non-invasive tier. Its differentiation (clinical-grade non-invasive comms for locked-in patients) is real, but the capital market has not funded it like a frontier BCI bet — it has funded it like an assistive-device / research-tools company, which is what the syndicate and the valuation say it is. No multiples are computable (no revenue) — EV/Sales, P/E: n/a.

Catalysts that move the (private) value (+private)

Events that have re-rated or could re-rate Cognixion:

• May 2023 — FDA Breakthrough Device designation. The single biggest credibility catalyst; unlocked the clinical narrative and likely the 2023–2024 financings.
• May 2024 — Series A (~$9.7M, per Tracxn). First "A" after years of seed/bridge — a milestone, but small.
• May 2025 — Blackrock Neurotech + BIOPAC distribution of Axon-R. Converted the platform into a sellable research product with third-party reach — the first genuine commercial catalyst.
• Jan 2025 — ALS longitudinal trial launch (NCT06810219, n≤10, runs ~through Apr 2026). The de-risking event the whole thesis hinges on.
• Oct 2023 / Jan 2026 / (reported) Feb 2026 — convertible-note raises. Survival financing; the pattern (debt over priced equity) is itself a (negative) signal.

Pattern: value here moves on regulatory milestones and distribution deals, not on a financial tape. The next real catalyst is trial readout (2026) → FDA submission path.

---

Phase C — Judge people & books

Management

• Andreas Forsland — Founder, CEO, Chief Product Officer. Designer/product leader (prior roles at/with Apple, IBM, Philips — Sr. Manager of Design, Citrix — Sr. Director Brand) who founded Smartstones (2013) then Cognixion (2014) after his mother's intubation left him as her interpreter — the authentic origin of the mission. Archetype: mission-driven founder-designer. Strengths: vision, product/UX, narrative, durability (12 years, kept it alive through bridges). Gaps: this is not a commercial-scale operator or a clinical/regulatory veteran profile — and the cap structure (serial seed + convertibles, sub-$50M after a decade) reflects a founder who can sustain and inspire better than he can scale capital or revenue.
• Other execs (lightly sourced, verify): CTO Chris Ullrich (neuroscience/engineering background) and COO James Williams are named via The Org; a prior search probe for "Brent Vasher COO" returned nothing — James Williams is the COO of record. Depth of bench beyond the founder is thin/undisclosed [n/a].
• Track record (quantified): the honest scorecard is one FDA Breakthrough designation, two distribution partnerships, one ongoing 10-patient trial, ~$25–32M raised over 12 years. That is real persistence and real regulatory traction — but no product clearance, no scaled revenue, and a predecessor company that closed.
• Capital allocation: capital-light and disciplined out of necessity (used grants/strategic money, partnered for distribution rather than building a sales force). No value-destroying M&A; equally, no value-compounding engine yet. ROE/ROIC: n/a — pre-revenue.
• Skin in the game / red flags: founder ownership presumably high after a seed-heavy history (undisclosed); no related-party, comp, or governance red flags surfaced. The only structural "flag" is single-point-of-failure key-man risk — this is Forsland's vision and standing with the ALS community as much as it is a company.

Forensic Red Flags + Regulatory

Forensic accounting: Not applicable in the usual sense — no audited statements, no 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes., no segment/SBC/goodwill disclosures to interrogate ``. The relevant "forensic" risks for a private at this stage are: (a) runway/going-concern — the convertible-note drip (2023/2026) is the tell to watch; (b) revenue quality — any "revenue" is small research-tool sales + grants, not product revenue, and should not be read as commercial traction; (c) trial-integrity — a sponsor-run, single-arm, n≤10 longitudinal study is hypothesis-generating, not pivotal — beware over-reading a glossy readout.

Regulatory findings (required sub-section):

• SEC (EDGAR LR + AAER): none possible — no CIK; private, not an SEC filer. total_sec_findings: 0.
• FDA: the only material regulatory fact is a positive one — FDA Breakthrough Device Designation (May 2023) for the BCI+AR assistive-communication device. No FDA clearance/approval yet; no recalls or warning letters surfaced.
• Non-SEC enforcement (FTC/DOJ/FDA/CFPB) web search: no enforcement actions, lawsuits, recalls, settlements, or consent decrees found for Cognixion or predecessor Smartstones. (Search noise returned unrelated "Cogni-" companies — Cognizant, Cognism, Cognitronics — none are this entity.)
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER; n/a for a private), web enforcement search, and public press as of 2026-06-21. The one regulatory event of note (FDA Breakthrough designation) is favorable. Smartstones being "permanently closed" is a business wind-down, not a legal/enforcement event.

---

Phase D — Project & stress-test

IPO-readiness & path-to-tradeable

No EPS model is possible (pre-revenue, private). The +private question is: how far is this from a tradeable event, and which event?

IPO-readiness: ~1–2 / 5 (early-to-growth; not late-stage, not pre-IPO) ``. Justification:

• Against an S-1: sub-$50M valuation, ~$25–32M lifetime raised, no product clearance, no scaled/recurring revenue, ~30 employees, no crossover investors, and reliance on convertible notes. None of the structural pre-IPO markers are present. An IPO is not the realistic exit.
• The realistic tradeable event is M&A. Cognixion is exactly the profile a larger player acquires for (a) the FDA-Breakthrough non-invasive comms IP, (b) the SSVEP+AR+LLM stack, (c) the ALS-community/clinical standing. The credible acquirers are visible in its own orbit: Blackrock Neurotech (already its distributor; wants a non-invasive line), an AAC incumbent (Tobii Dynavox extending beyond eye-tracking into late-stage patients it currently can't serve), a big-tech accessibility/AR group (Apple — there's already a Vision Pro integration; Amazon — already an investor via Alexa Fund; Snap — already bought NextMind), or a medtech distributor..

Milestones that would unlock a step-up (or an acquisition bid):

1. Positive ALS trial readout (2026) showing reliable, faster-than-eye-tracking communication in locked-in patients → the core de-risk.
2. An FDA submission/clearance pathway post-Breakthrough → turns "candidate" into "clearable product."
3. A reimbursement path (Medicare DME-equivalent, as eye-trackers enjoy) → turns clearance into revenue.
4. Axon-R research revenue scaling through Blackrock/BIOPAC → proves a real near-term P&L and funds the wait.

Write-back: there is no cognixion entry in research/private-watch.json (only neuralink, synchron). Recommend adding one — beat: bci, stage: growth, ipo_readiness: 1, lead_investors: "Prime Movers Lab, Amazon Alexa Fund, Northwell", catalyst: "ALS trial readout 2026 → FDA path; M&A-likely exit", dossier: this file. (Per wave rules I am not editing watch/registry files in this run — flagging for the conversational queue step.)

Bull vs Bear

Bull case. Cognixion owns the only non-invasive, surgery-free, AR-native communication wedge for the patients implant players won't reach for years and eye-trackers physically can't serve once ocular motility goes. An FDA Breakthrough designation, a 2026 trial, two distribution partners (one of them an implant leader that chose to carry the non-invasive line), and a per-patient LLM co-pilot that gets better as foundation models do — all for a sub-$50M entry price. If the trial reads out well, this is a cheap, high-optionality acquisition target for Blackrock, Tobii, Apple, Amazon, or Snap, where a single strategic bid could be a multiple of the last mark. The secular tailwind (BCI legitimacy, generative-AI decoding, an aging ALS/stroke population, >4M Americans on AAC) lifts the whole category.

Bear case (permanent-impairment risks).

1. It never becomes a product, only a project. Twelve years, sub-$50M, convertible-note bridges, no clearance, no scaled revenue — the base rate for an assistive-tech company stuck in this loop is a quiet wind-down or a fire-sale acqui-hire (its own predecessor, Smartstones, closed). The clinical/reimbursement gauntlet is long and underfunded relative to the burn.
2. Surrounded and out-capitalized. Non-invasive consumer/research players (Emotiv, OpenBCI, Neurable, Snap/NextMind) compress the research-tools side; invasive leaders (Neuralink, Synchron, Precision — with 10–100× the capital) will, in time, push down into the same locked-in comms use-case from the high end. Cognixion is squeezed from both directions.
3. Reimbursement/efficacy gap. SSVEP non-invasive BCIs are slower and noisier than implants; if real-world communication speed doesn't decisively beat existing eye-tracking SGDs for the patients who still have eye control, the addressable population shrinks to the truly locked-in — a small, hard-to-reimburse niche.

Pre-mortem (18 months out, thesis broke): the n≤10 trial read out fine-but-not-decisive (works, but not clearly faster/more reliable than incumbents for most patients), FDA clearance stayed years away, Axon-R research sales stayed in the low hundreds of units, the convertible holders converted at a flat/down mark, and the company was absorbed in a modest acqui-hire — or simply ran the runwayRunwayHow long the cash lasts at the current rate of spending. It shortens the moment spending rises, which is why a figure taken from a quiet quarter flatters. down.

Are the marks too high? No — at ~$10–46M it is not expensively priced; the risk isn't overvaluation, it's terminal-value/exit uncertainty (does it ever clear + reimburse + scale, or get bought cheaply).

Contrarian view of what the market refuses to see: Cognixion should arguably be valued as an AI-assistive-communication + research-tools company with an embedded acquisition put, not benchmarked against frontier BCI unicorns. Framed that way, the sub-$50M mark plus a 2026 catalyst plus named strategic acquirers in its own orbit is a reasonable small-cap-private asymmetry — the upside is a strategic buyout, not an IPO, and that's fine.

Devil's Advocate (short-seller)

Dismantling the bull case:

• The revenue is a mirage. "Now available," "commercial," "B2B subscriptions" describe a $25K research headset sold in limited quantity through other companies' sales forces. There is no disclosed product revenue, no recurring base, no scale. Treat any "traction" language as marketing, not a P&L.
• Concentration = the trial. Essentially 100% of the equity story rests on one sponsor-run, single-arm, ≤10-patient study. That is not a pivotal trial; it cannot, by design, clear the device. The most likely outcome is a positive-but-soft readout that still leaves years and millions of dollars between here and clearance + reimbursement.
• The moat is thinner than it looks. Non-invasive EEG/SSVEP is a commoditizing modality (Emotiv, OpenBCI, g.tec sell the hardware; the AR layer is increasingly Apple/Snap's to provide; the LLM is a third-party API). Strip those out and what's proprietary is an integration + a per-patient fine-tune — defensible for a niche, not a fortress.
• Most dangerous competitor bulls underrate: Tobii Dynavox, not Neuralink. Tobii already owns the ALS communication customer, the clinician relationships, and Medicare DME reimbursement. If Tobii bolts a generative-AI co-pilot and a basic SSVEP fallback onto its installed base, Cognixion's wedge (the late-stage hand-off when eye control fails) gets absorbed by the incumbent for a fraction of the effort.
• Capital-allocation / structure flags: the convertible-note drip (2023, 2026, reported Feb 2026) is what companies do when a priced round is hard to raise — it caps dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company. optics now at the cost of an overhang later, and it signals investor reluctance at a clean valuation.
• What must hold for today's ~$10–46M mark: that the 2026 readout is clearly superior to eye-tracking, that an FDA pathway materializes, that reimbursement follows, and that a strategic acquirer pays up before the cash runs out. Miss the trial bar by 20–30% (i.e. comms speed/reliability merely matches incumbents) and the realistic outcome is a flat-to-down convertible conversion and a modest acqui-hire.
• Single scenario that permanently impairs it: trial reads out underwhelming + FDA timeline extends + Axon-R sales stall → runway exhausts before the next bridge → wind-down or fire-sale. Plausibility: moderate-to-high given the predecessor's fate and the financing pattern.

Management Questions (ordered by information value)

1. What were the quantitative results of the NCT06810219 ALS trial — words-per-minute, selection accuracy, and reliability vs. each patient's prior eye-tracking SGD — and in how many of the ≤10 patients did Axon-R outperform their existing device?
2. What is your specific FDA pathway post-Breakthrough (De Novo? 510(k) predicate? PMA?), and what is the realistic clearance timeline and remaining clinical spend?
3. What is current cash, monthly burn, and runway, and why the recent reliance on convertible notes rather than a priced equity round?
4. What is the reimbursement strategy — will Cognixion ONE Axon qualify for Medicare DME like eye-tracking SGDs, and what's the evidence threshold?
5. How many Axon-R units have actually shipped through BIOPAC and Blackrock, at what ASP, and what is 2025/2026 research-tools revenue?
6. The distribution deals are non-exclusive — what stops Blackrock or BIOPAC from dropping the line, and do you have any path to owning the customer relationship?
7. Why should we believe Tobii Dynavox (or another AAC incumbent) can't replicate your late-stage hand-off by adding an AI co-pilot + SSVEP fallback to its reimbursed installed base?
8. What is proprietary and patented in the SSVEP + AR + per-patient-LLM stack, versus integration of third-party EEG hardware, AR optics, and foundation-model APIs?
9. Who is the foundation-model provider for the conversational co-pilot, and what's your exposure to its pricing, availability, and data-privacy terms (PHI)?
10. What does the bench beyond you look like — clinical/regulatory and commercial leadership — and what is the key-man succession plan?
11. What is the realistic exit thesis — clear-and-scale independently, or be acquired — and which strategic buyers are you building toward?
12. How do you compete for the same locked-in patient against invasive players (Synchron, Neuralink) as their implants mature and reimbursement arrives?
13. What is the cap-table structure today — founder/employee vs. investor ownership, convertible-note caps/discounts outstanding, and liquidation preferences?
14. What did you learn from Smartstones' closure, and what is structurally different this time?
15. What is the gross margin profile of Axon-R hardware, and can research-tools revenue realistically fund the company to FDA clearance without further dilution?

---