Private AI for Alberta Psychologists — Session Notes Stay Local

Before: the documentation burden

Dr. Anika Chen runs a private practice in Calgary, seeing seven clients per day. She is present, engaged, and thorough in every session. And then the sessions end.

After the last client leaves, the documentation begins. Progress notes for each session — what was discussed, what was observed, what was planned. For a 50-minute session, a complete DAP note (Data, Assessment, Plan) typically takes 15 to 20 minutes to write. Surveys of mental health professionals indicate 8 to 15 minutes per note is average, with complex cases taking up to 30 minutes (ICANotes, 2021).

Seven clients. 15–20 minutes each. That is 90 to 140 minutes per day spent on documentation that, however necessary, does not help a single client.

Dr. Chen has looked at AI note-writing tools. Several exist. But each one has the same architectural problem: session notes contain protected health information, identifying details, clinical observations. Sending that information to a cloud server — even one with a privacy policy — carries real risk under Alberta's PIPA and the standards of practice set by the College of Alberta Psychologists (CAP).

CAP's September 2024 guidance is clear: psychologists bear full professional responsibility for their records regardless of what tools they use to create them. The obligation to protect client confidentiality does not transfer to a software vendor. Dr. Chen is responsible for her notes, which means she is responsible for where they go.

After: documentation that fits the workday

With Ferox Nodus running on her Mac, the note-writing process looks different. After each client leaves, she opens Ferox, selects a session note template, and types her clinical observations. The AI generates a structured DAP note draft — on her device, using Google's Gemma 3 — and nothing leaves her machine.

She reviews the draft. She edits, adjusts, and adds her clinical judgment. She approves and saves. The whole process takes five to eight minutes instead of fifteen to twenty.

A systematic review found AI-assisted clinical documentation reduced documentation workload with a standardized mean difference of −0.71 — a meaningful reduction with direct implications for practitioner wellbeing (Zhao et al., BMC Medical Informatics and Decision Making, 2025).

Across seven clients per day, five days per week, the time savings compound. Documentation goes from the long tail of a full workday to a manageable close.

Bridge: the practice she wants to run

Ferox Nodus is the bridge between the documentation burden Dr. Chen has and the practice she wants to run — one where the end of the session is the end of the clinical day.

The AI does not replace her judgment. It generates a structured draft that she reviews, edits, and approves. Her clinical assessment, her professional signature, her responsibility — she does not abdicate any of that. She simply spends less time on the mechanical task of structuring her observations into a compliance-ready format.

For non-clinical tasks — research, continuing education summaries, administrative correspondence — Ferox can use cloud AI with PII scrubbing enabled. Only clinical session notes stay fully local. Zero telemetry. Hardware-level encryption. Runs on your device.

A controlled experiment found professionals using AI assistance completed tasks 25% faster and produced 40% higher quality outputs compared to working without AI (Harvard Business School and Boston Consulting Group, 2023). For documentation that has to be accurate and complete, that quality signal matters.