Nabla

What Nabla actually does

Nabla — sometimes shipped as "Nabla Copilot" — is an ambient AI clinician assistant. The clinician opens the copilot, starts a recording, and Nabla handles transcription, speaker separation, and structured note generation in the EHR's expected format. The product surfaces the company is investing in for 2026 are coding, deeper EHR integration, and "agentic" workflows that close out items from the encounter (orders, follow-ups, referrals) rather than just drafting the note.

Deployment posture

Cloud-only commercial SaaS, but with the most aggressive data-minimization defaults among the major commercial scribes. The standard configuration does not retain audio, retains structured medical data for 14 days by default, and does not train on customer data. That is not the same as "on-prem" — the inference still happens in Nabla's cloud — but it is the closest a cloud vendor in this category gets to a "least data possible leaves the building" posture.

The evidence base

Nabla has the strongest published clinical evidence of the commercial scribes profiled in this directory:

• checkUCLA randomized clinical trial (NEJM AI, 2025) — three-arm pragmatic RCT, 238 outpatient physicians, 14 specialties, randomized 1:1:1 to Nabla, Microsoft DAX Copilot, or usual care, November 4, 2024 – January 3, 2025. Nabla physicians saw an estimated 41-second reduction per note (4:30 → 3:49), a statistically significant 9.5% drop versus control. DAX showed a smaller decrease that did not reach statistical significance. Both arms reported ~7% burnout improvement. Trial registry: NCT06792890.
• checkScale evidence — Nabla has now generated more than 8 million AI clinical notes across its customer base. The 85,000-clinician footprint is one of the largest in the category, second only to Abridge.
• closeSafety caveat from the same RCT — UCLA investigators noted that AI-generated notes occasionally contained clinically significant inaccuracies, most often omissions or pronoun errors. One mild patient-safety event was reported during the study window. The signal-to-noise of any ambient scribe still requires clinician review; Nabla's evidence is not an exception to this.

Privacy and compliance posture

Among the major commercial scribes, Nabla has the most defensible privacy posture out of the box. Default settings: no audio storage, 14-day medical data retention, no training on customer data. The compliance stack is wider than U.S.-only peers — HIPAA + GDPR + ISO 27001 + SOC 2 Type II — reflecting the company's European founding and dual-market posture.

The defaults are configurable, which is the right place to focus during contracting. A customer that wants to keep the no-audio-by-default and 14-day-by-default behavior should write those numbers into the contract; a vendor's published defaults can be silently changed if not contractually fixed. For a buyer evaluating Nabla against its peers, this is an unusual dimension on which to be doing the homework — and a real advantage.

That said, Nabla is still a cloud product. Audio is processed off-prem; data is processed off-prem. For Canadian provincial settings under PHIPA, HIA, or Quebec's Law 25 with strict residency interpretations — or any buyer whose binding constraint is "PHI does not leave the building" — Nabla's defaults reduce the surface area but do not eliminate it. The on-prem path is still the cleaner answer for that specific constraint.

Strengths and limitations, plainly

Where Nabla fits versus an on-prem alternative

If the framing is "we want the best-evidenced cloud scribe with the cleanest privacy defaults," Nabla is the answer. If the framing is "audio cannot leave our building under any configuration," the on-prem path is still the cleaner one — Nabla's defaults reduce the surface area but do not eliminate it.

Where this fits in the WalledCare directory

Nabla is the strongest commercial reference for the evidence-led framing of AI Scribes. If your team's evaluation rubric weights peer-reviewed RCTs, multilingual coverage, or default privacy posture, Nabla deserves to be on the shortlist alongside Abridge, Ambience, and DeepScribe. If the framing is on-prem residency, read our on-prem reference stack first.

Quick facts

Use Nabla as the benchmark, then decide

If the deciding factor is "strongest published evidence and the most conservative privacy defaults among cloud scribes," Nabla is the shortlist anchor. If the deciding factor is "audio and data cannot leave our network at all," WalledCare's on-prem stack is the alternative built for that.

send Request a WalledCare pilot arrow_back All vendors