Piramidal

What do they actually do

Piramidal builds a large AI model trained on EEG (brainwave) data and is applying it first as a clinician copilot for epilepsy diagnosis. The product ingests EEG recordings from different machines, harmonizes the data, and generates time‑stamped alerts and standardized summaries so clinicians can review hours of signals in seconds rather than manually scanning long traces YC, TechCrunch.

Today the company is focused on hospital pilots, especially in ICUs, where teams must monitor continuous EEG and act quickly on evolving brain events. Reported signals include a $6M seed round and planned multi‑site ICU pilots to validate performance, workflow fit, and safety before broader rollout TechCrunch, YC. Later reporting describes ICU deployments and partnerships as the team moves from demos into real‑world monitoring pilots Wired, AHA.

Who are their target customer(s)

• ICU teams (critical-care physicians and nurses): They must continuously watch EEG feeds for fast-changing events, but reviewing long recordings by hand is slow and easy to miss, delaying treatment. They need reliable alerts and concise summaries to act faster and reduce manual review burden (TechCrunch, Wired).
• Hospital neurologists and epileptologists: Heavy caseloads and device-to-device variability lead to backlogs and inconsistent reads. Standardized, rapid readouts help prioritize urgent cases and reduce variability across readers and machines (YC, TechCrunch).
• Outpatient EEG clinics and EEG technologists: Ambulatory and overnight studies require many hours of manual scoring, creating technician load and long turnaround times. Automating the initial pass can cut tedious review without changing clinic workflows (YC).
• Clinical-trial and research teams using EEG endpoints: Multi-site trials struggle with consistent, reproducible EEG biomarkers across different devices and sites. A harmonized model can standardize endpoints and speed analysis to make multi-center biomarker work feasible (YC, TechCrunch).
• Remote-monitoring and device integrators (home EEG/telemetry): They collect large volumes of continuous EEG but lack a scalable, accurate analysis layer that works across varied electrode setups and minimizes false alarms for remote care workflows (YC).

How would they acquire their first 10, 50, and 100 customers

• First 10: Use warm intros from YC, investors, and clinical advisors to run hands-on ICU and epilepsy-center pilots in ~10 hospitals. Offer tightly scoped, no-charge co‑development in exchange for anonymized EEG data and feedback; capture standardized KPIs and publish de‑identified case studies to convert pilots into paid references (YC, TechCrunch).
• First 50: Leverage pilot case studies and clinician co‑authors to present at neurology/ICU conferences and in journals; convert referrals from early sites and integrate with common EEG hardware/software vendors to ease deployment. Add a small clinical sales team to close multi‑site neurology and ambulatory-monitoring groups (YC, TechCrunch).
• First 100: Package a repeatable implementation playbook (configs, compliance, training, SLAs) and sell health-system contracts plus standardized bundles for outpatient and remote-monitoring vendors. Open channel partnerships with EEG telemetry/telehealth platforms and offer validated EEG endpoint analytics to sponsors/CROs (YC, TechCrunch).

What is the rough total addressable market

Top-down context:

Near-term, EEG analysis software is commonly estimated in the ~$0.5–0.7B range (e.g., ~$500M in 2023; ~$653M in 2024) DataIntelo, MarketGrowthReports. Broader EEG devices/systems spend is ~USD 1.2B in 2023 with growth through 2030, to which analysis layers attach Grand View Research. Adding related segments (ambulatory EEG; brain monitoring) pushes the platform opportunity into multi‑billion territory over time DataIntelo ambulatory EEG, MarketsandMarkets brain monitoring.

Bottom-up calculation:

Clinically, the U.S. performs >1M EEGs/year, including ~250k continuous EEG studies in ICUs; at $50–$100 per analyzed study, that’s ~$50–$100M/year in the U.S. alone, before ambulatory and home telemetry volumes. Extending similar pricing across multi‑million global studies supports a software TAM on the order of several hundred million dollars, consistent with top‑down ranges (Epilepsia HEEDB).

Assumptions:

• Per‑study analysis pricing averages $50–$100 for routine/ambulatory studies; higher for cEEG bundles.
• A subset of EEG exams is addressable initially (hospital and ambulatory settings), expanding with validated performance and integrations.
• Top‑down device and software figures overlap in some reports; analysis TAM focuses on software/service revenue layered on existing EEG infrastructure.

Who are some of their notable competitors

• Ceribell: Point‑of‑care EEG system plus Clarity cloud algorithm for real‑time seizure alerts in ICU settings; hardware+software with published outcomes and FDA 510(k) clearance for its seizure‑detection algorithm (product, FDA press).
• Persyst: Widely used EEG review suite with automated detectors (spike/seizure) integrated into many hospital systems; directly overlaps on automated surfacing of suspicious segments and is embedded in clinical workflows (overview, comparison study).
• Natus (NeuroWorks / autoSCORE / BrainWatch): Established EEG device/software vendor adding AI features (autoSCORE) and a point‑of‑care BrainWatch system; strong enterprise footprint and turnkey integrations (autoSCORE, BrainWatch).
• encevis: European EEG viewer with validated automated modules (seizure, spike, artifact). Competes on detector accuracy and lightweight integration for neurophysiologists (product, clinical eval).
• Zeto: Cloud‑first EEG platform for ambulatory/clinic workflows, adding AI seizure‑burden software (NeuroPulse) and partnerships for automated analysis—overlaps with remote/tele‑EEG use cases (announcement).