About ConfideAI
Why I Built This
ConfideAI began with a lunch conversation. A friend in clinical practice mentioned that her service couldn't use AI because of confidentiality concerns, then described the sheer volume of notes, reports, and letters that were swallowing her week.
I'm trained as a psychologist, though my career has leaned more toward the organisational side — understanding how people, workflows, and technology fit together. From that vantage point, her problem looked solvable, but only if AI could be introduced in a way that preserved strict clinical confidentiality while taking a meaningful chunk out of the documentation burden.
When large language models arrived, I thought the problem was finally solved. AI could draft a SOAP note in seconds. It could structure a treatment plan from session notes. The technology was ready.
But the ethics weren't.
Every AI tool I looked at required uploading clinical data to a cloud server. That meant trusting a third party with your clients' most intimate disclosures — their trauma histories, their diagnoses, their words spoken in confidence. The cloud provider could see it. Their employees could access it. A breach could expose it. And the clinician using the tool had no way to verify otherwise.
Most therapists I spoke to reached the same conclusion I did: the risk wasn't worth it. So they kept doing everything by hand.
That's the gap ConfideAI was built to close.
The Problem No One Was Solving
The mental health profession has a documentation problem that runs deep. Research consistently shows that clinicians spend significant portions of their working week on administrative tasks — writing notes, updating plans, drafting correspondence. That's time taken directly away from the people who need help.
AI-powered documentation tools exist, and some of them are genuinely good at generating clinical text. But they all share the same fundamental flaw: they rely on trust.
Trust that the provider won't read your data. Trust that their security won't be breached. Trust that their employees won't have access. Trust that their policies won't change.
For mental health practitioners, trust-based privacy isn't good enough. Confidentiality is the foundation of the therapeutic relationship. It's not a feature — it's the ethical bedrock of the entire profession.
What Makes ConfideAI Different
ConfideAI doesn't ask you to trust us with your clinical data. ConfideAI is designed to minimise how much trust you have to place in any provider, by enforcing confidentiality at the hardware and architectural level.
The platform runs on confidential computing infrastructure — specifically Intel TDX processors and NVIDIA H100 GPUs with hardware-level security. When you send clinical content to ConfideAI for processing, it enters a Trusted Execution Environment — a hardware-secured enclave where data is protected even while it's being processed. Not before. Not after. During.
ConfideAI does not store any clinical content on its own servers. Data travels over TLS 1.3 encrypted connections and is processed inside TEE-protected enclaves. Learn more on our Security & Privacy page.
Built for Clinicians, Not IT Departments
ConfideAI was designed from the ground up for the way mental health practitioners actually work — not for enterprise procurement teams or hospital IT departments.
The platform includes 20 clinical templates covering the full documentation lifecycle: intake assessments, SOAP, BIRP, and DAP progress notes, treatment plans with SMART goals, psychoeducation handouts, discharge summaries, referral letters, and more. Every template is grounded in clinical literature, with built-in support for over fifteen therapeutic modalities including CBT, ACT, DBT, psychodynamic, Jungian, person-centred, and narrative approaches.
One of the features practitioners tell us they value most is document chaining. The output from one template feeds directly into the next, creating a connected workflow from intake through to discharge. You bring your clinical judgement, the referral documents, and your ongoing session and assessment notes. ConfideAI can handle the rest of the chain.
Clinical drafts ready for your review and refinement.
Our Technology Platform
ConfideAI runs on NEAR AI's confidential computing infrastructure, built with Intel TDX processors and NVIDIA H100 GPUs. Intel's TDX technology creates secure processing enclaves, and NVIDIA's H100 GPUs provide the computing power for AI inference inside those enclaves.
Where We Are Today
ConfideAI is currently in closed beta with real practitioners using the platform for their clinical documentation. We're based in Australia, initially focused on Australian mental health professionals, with plans to expand globally as we grow.
The platform is free during the beta period. We believe strongly that AI-powered clinical documentation should be affordable for independent practitioners and solo clinicians — not just large practices with enterprise budgets. Our pricing model, when it launches, will reflect that commitment.
If you're a mental health practitioner interested in joining the beta, reach out to us at support@confideai.ai.
What We Believe
Privacy is non-negotiable. Not "privacy-first." Not "privacy-focused." Non-negotiable. We built the entire platform around the principle that clinical data must remain confidential — enforced by hardware-level isolation during AI processing and an architecture designed to minimise data exposure at every layer.
Built for clinicians, not corporations. Every design decision starts with one question: does this help a practitioner spend less time on paperwork and more time with clients? If the answer is no, we don't build it.
Affordable for independent practitioners. The therapists who need this most are often solo practitioners running their own practice. ConfideAI will always be priced so that independent clinicians can access it.
Your data stays under your control. ConfideAI doesn't retain or monetise clinical content, and we don't train on your notes. When infrastructure providers handle your data, it remains encrypted at rest and is processed inside secure hardware enclaves, accessible only via your authorised login.
The Mission
ConfideAI exists to make AI-powered clinical documentation accessible to every mental health practitioner, without compromising the ethical standards that define the profession.
Therapists chose this career to help people, not to spend their evenings writing notes. If we can give them those hours back — safely, ethically, and affordably — then we've done what we set out to do.
Want to learn more or join the beta? Contact us at support@confideai.ai