BCI Atlas

Paradromics’ Connexus BCI: high-channel-count intracortical recording aimed at restoring speech and computer control, plus what’s known publicly about the company.

Company — American

Paradromics (company brief)

BCI · implant · intracortical · speech · Paradromics · Connexus

Paradromics is building an implantable, high data‑rate brain–computer interface (BCI) designed to record activity at (or near) the single‑neuron level and translate it into speech/text and computer control. Their flagship clinical product is the Connexus® BCI, aimed first at restoring communication for people with severe motor impairment (e.g., ALS, spinal cord injury, stroke).

This brief focuses on the technology, clinical path, and company facts we can support with public sources.

At a glance

Funding (publicly stated / reported)

The technology (Connexus BCI)

Paradromics describes Connexus as a fully implantable system intended for long‑term use, using materials common in medical implants (e.g., titanium alloy housing, platinum‑iridium electrodes) and miniaturized microelectrodes.

1) Implant architecture (as publicly described)

Paradromics describes three main implanted components:

  • Cortical module (implanted under the skin; microelectrodes extend just below the brain surface)
  • Internal transceiver (implanted in the chest)
  • Extension lead (connects cortical module to chest transceiver)

They describe the chest transceiver as handling wireless optical data transmission to an external transceiver worn by the user, and inductive power transfer (wireless charging style) to power the implant.

2) Channel counts + scaling strategy

Paradromics claims:

3) Decoding stack (software)

Paradromics describes sending data to a small computer running AI / machine learning / language models to translate neural signals into:

Clinical path (what they say is happening)

IDE approval + early feasibility study (Connect‑One)

Paradromics states the U.S. FDA granted Investigational Device Exemption (IDE) approval to begin the Connect‑One Early Feasibility Study (EFS) with Connexus.

They state initial enrollment begins with two participants near three clinical sites:

  • UC Davis
  • Massachusetts General Hospital
  • University of Michigan

(Details and eligibility criteria are on their clinical study page.)

What they’re doing recently (publicly visible milestones)

From Paradromics’ own announcements:

How Paradromics fits in the BCI landscape (my framing)

Paradromics looks like a classic “high‑bandwidth intracortical” bet:

  • maximize information throughput by recording as close to neurons as possible,
  • keep the device fully implantable and durable,
  • focus first on a clinically defensible indication: assistive communication.

Key questions to watch

  1. Signal stability over time
  • Chronic single‑unit / high‑density interfaces face long‑term encapsulation and stability issues.
  1. Surgical workflow + explant/upgrade path
  • How invasive is implantation, and what happens when parts fail or need replacement?
  1. True communication performance
  • The meaningful metric is real‑world, daily use: words per minute, error rates, retraining burden, latency.
  1. System bottlenecks
  • Power budget, heat, data link reliability, and robust decoding under realistic conditions.

Notes on sourcing

  • I prioritized Paradromics’ own pages for claims about device architecture and study plans.
  • For funding/BD designation, I included both the company announcement and third‑party coverage.
  • If you want employee count / headcount estimates, we can add LinkedIn data when it’s reliably accessible in our tooling.