Cycling through all six. Tap any point to stop.

Measured on six things

AI-powered real-time SOAP note generation for physical th...

Growth

Young company (launched ~2025) with PT-heavy team. Building custom LLM tailored to physical therapy. Early stage with Notation as flagship product. Growing interest in PT AI documentation space but faces stiff competition from established EMRs and other scribes.

Screenshots

1 / 4

FOWND

About this app

FOWND is a game-changer for physical therapists drowning in paperwork. It uses AI to listen in on your sessions and automatically generate compliant SOAP notes in real time. No more staying late to write up notes or sacrificing patient interaction for documentation. It's built specifically for PT workflows, integrating seamlessly into your practice. The result? Less admin time, more time with patients, and improved care quality. I've seen too many burnout cases from documentation overload; FOWND is the relief we've been waiting for.

Key features

1Real-time ambient audio capture that generates compliant SOAP notes during the session
2Custom PT-specific large language model (not a generic wrapper) for better clinical accuracy
3Computer vision analysis for movement, posture, gait, and functional assessment documentation

Target customer profile

Drawn from the product itself, not from a survey.

Demographic

Independent physical therapists in private practice

Pain points

Spending hours after work on documentation; struggling to keep up with insurance compliance.

Primary needs

Reduce documentation time; accurate, compliant notes; more time for patients.

Demographic

Large PT clinic chains with multiple therapists

Pain points

Inconsistent note quality across staff; high admin overhead; therapist burnout.

Primary needs

Standardized documentation; improved efficiency; reduce turnover.

Demographic

Traveling physical therapists

Pain points

Need to document quickly in varying environments; repetitive note-taking across different settings.

Primary needs

Easy-to-use mobile solution; rapid note generation; compliance across states.

Scout Review

Written by AI from measured evidence, scored out of 100.

FOWND/Notation is one of the more promising PT-specific AI documentation tools because it's not a generic medical AI repackaged — it's built by PTs with a custom model and unique computer vision capabilities. The real-time ambient listening approach could genuinely transform workflows for independent and traveling therapists drowning in paperwork. It outperforms template-based alternatives in automation and should reduce burnout. That said, it's early stage, public information is limited, and like all AI scribes it will require human oversight. Solid product addressing a real crisis in our profession.

Measure by measure

Usability

Real-time SOAP generation from session audio, purpose-built for PT workflows (evals, assessments, plans). Mobile-friendly for traveling therapists. Minimal manual input compared to template tools. Editing required post-generation like most AI scribes.

Design

Clean, modern PT-focused interface with real-time note preview during sessions. Uses ambient audio capture and has computer vision elements for movement analysis. Professional healthcare aesthetic but limited public screenshots make full assessment difficult.

Speed

Live, real-time compliant SOAP notes generated during the session. Ambient voice recording eliminates post-session charting. Custom PT-specific LLM (not an OpenAI wrapper) helps with faster, more relevant output.

Security

HIPAA-compliant AI EMR. Built by PTs for PTs with emphasis on privacy for clinical audio. Focused on insurance compliance and defensible notes across states. No major red flags in available info.

Accessibility

Strong for traveling and independent PTs with mobile solution. Supports varying environments and multi-state compliance. Less ideal for large clinic chains needing deep EMR integrations. Limited info on WCAG or assistive tech support.

Growth

Improvement areas

More public transparency around accuracy rates, editing time required, and real user case studies
Clear integration details with popular PT EMR platforms and multi-user clinic features
Detailed pricing tiers and specific support for large clinic chains vs independent practitioners

Design & Usability

FOWND (specifically their Notation tool) stands out by being built from the ground up for physical therapists rather than bolting PT features onto a generic medical scribe. The combination of ambient audio listening and computer vision for analyzing gait, posture, and movement is genuinely innovative for our field. It promises to generate compliant SOAP notes in real time so therapists can stay present with patients instead of thinking about documentation. Compared to template-heavy tools, this reduces cognitive load significantly. Usability seems high for independent and traveling PTs, though large clinics may need more robust multi-user and admin features. The interface appears clean and focused, avoiding the cluttered feel of many EMRs.

Speed & Security

The core value proposition — real-time note generation during sessions — directly attacks the biggest pain point in PT: spending 1-2 hours after clinic doing notes. Their custom PT-specific large language model (explicitly not just an OpenAI wrapper) should deliver better accuracy on terminology, exercises, and functional outcomes than generic tools. Security is appropriately prioritized with HIPAA compliance for handling sensitive clinical audio. Speed appears excellent based on claims, though real-world performance will depend on audio quality in noisy clinic environments and how much editing is still required. Early feedback in the PT community on similar tools shows they save significant time but rarely eliminate review entirely.

Accessibility & Growth

Particularly strong for independent practitioners and traveling therapists who need flexible, mobile-first solutions that work across different settings and state regulations. Less information available on deep integrations with popular PT EMRs, which could limit adoption in larger clinic chains. As a young company founded by PTs, they understand the burnout problem intimately. The space for AI documentation in rehab is growing rapidly, and having a purpose-built tool with computer vision elements gives them a differentiated position. However, they need to prove accuracy and reliability at scale before widespread adoption. Growth will depend on delivering on the 'less admin time, more patient time' promise without introducing new headaches.

Conclusion

If you're a physical therapist tired of charting until 8pm, FOWND is worth trying. The combination of real-time generation, PT-specific training, and focus on compliance makes it a strong contender. Just don't expect it to be perfect on day one — review your notes carefully. This is the direction the industry needs to go to fight burnout and keep talented therapists in the profession.

Suggestions for improvement

Provide more transparent accuracy metrics and real PT case studies on the website
Clarify exact integrations with popular EMR systems like WebPT, Cliniko or others
Add clearer pricing information and detailed feature comparison for different practice sizes
Include video demos showing real session-to-note workflow with computer vision elements

How we score

Alternatives comparison

Named competitors, point by point. Nobody paid to appear here or to be left out.

	FOWND	SOAPnote.io	WriteUpp	Healigo
Real-time AI generation from ambient audio	Yes - live during session with custom PT LLM	No - template-based generator, manual input	No - traditional note-taking in clinic management	No - manual input with note templates
Computer vision for movement analysis	Yes - analyzes gait, posture, movement	No	No	No
Built specifically for PT workflows	Yes - by PTs, custom PT model, Notation tool	Partial - general SOAP templates	Partial - clinic management with notes	Yes - PT-specific EMR but manual
Documentation time reduction	High - real-time during session	Medium - speeds template filling	Medium - streamlines clinic processes	Low - requires manual entry