I’m a Licensed Clinical Social Worker with more than two decades of experience in psychotherapy, emergency psychiatric evaluation, crisis and safety work, multidisciplinary care, and clinical supervision. Since 2026, I’ve also been doing independent work on how AI can support clinicians without compromising privacy, safety, or clinical judgment, including Therapist Session Compass, a zero-PHI-by-design clinician-guidance concept.
My work has centered on assessment, treatment, risk, supervision, and coordinated care across outpatient, emergency, substance-use, and inpatient settings.
Psychosocial assessment, psychotherapy, treatment planning, crisis intervention, multidisciplinary care, chart review, critical-incident follow-up, and supervision of interns and clinical staff.
Hospital emergency-department psychiatric evaluations, suicide and safety risk assessment, disposition decisions, safety planning, and coordination with hospitals and law enforcement.
Behavioral therapy and medical social work, substance-use treatment, and inpatient psychiatric work, including Section 65 supervision and earlier hospital-based clinical experience.
For approximately three years, I worked in emergency departments where decisions had to be made quickly, often with incomplete information, and where safety, disposition, and clear handoff mattered immediately.
This work required integrating presentation, history, stressors, protective factors, collateral information, and immediate risk indicators into decisions that had real consequences for patients, families, and care teams. Good disposition work depended on knowing what was known, what remained uncertain, and what level of response the situation actually required.
Beyond direct treatment, my work has included supervising interns and clinical staff, program-level supervision, chart review, critical-incident follow-up, and coordination across multidisciplinary teams.
Supervision of interns and clinical staff across multiple organizations, with attention to clinical reasoning, documentation, professional development, and safe decision-making.
Chart review, critical-incident review and follow-up, board participation, and collaboration across disciplines when care, safety, and organizational responsibility intersect.
Years of working with risk, uncertainty, documentation, and high-stakes clinical decisions left me curious about how those same problems show up when AI gets introduced into clinical work. In 2026, that curiosity became an independent project: using my clinical background to test where AI can responsibly support a clinician, and where it cannot.
These examples show the same method from two directions: build something useful for clinicians, then make its boundaries and failure modes inspectable.
Problem: clinician-facing AI can become either too generic to help in the room or too confident about decisions that belong to the clinician. Approach: Session Compass is a zero-PHI-by-design concept that translates curated psychotherapy evidence into structured in-session guidance — interventions, questions, exercises, psychoeducation, and cautions — without storing a client record. Twenty-four underlying evidence sources were consolidated into seventeen retrieval-oriented modules with explicit scope rules, source lists, supported points, boundaries, product rules, and card mappings. Condition-specific guidance remains separate where merging it would reduce retrieval precision. Constraint: the system surfaces bounded evidence and options; clinical judgment and authority remain with the clinician.
Problem: high-risk requests could still produce unsafe clinical-support behavior despite instruction-level safeguards. Method: structured testing across crisis timing, safety-clearance language, PHI, diagnosis, and treatment boundaries. Result: the workflow moved toward bounded task modes, checks before and after model output, and fail-closed handling when intent is unclear.
Problem: “unsafe” is too vague to test consistently. Method: convert clinical risk into explicit categories with expected routing behavior, then test both missed risk and unnecessary blocking. Result: a nine-category taxonomy and a self-authored 34-case evaluation suite with eight negative controls.
The core question I’m testing: how can AI support a clinician without quietly taking over decisions, privacy, or authority that should stay human? Therapist Session Compass, a zero-PHI-by-design clinician-guidance concept, is the clearest example of that work so far.
Rebuilt from first principles for a single authorized clinician. It separates general and clinical workspaces, keeps client context out of shared knowledge stores, and reserves diagnosis, treatment, crisis disposition, and record submission for the clinician, never the AI.
The design isolates each client/session, runs models locally, and keeps external transmission off by default. Reference material is tracked back to its source so suggestions can be traced to the material that informed them.
I test the system against a self-authored set of risk categories and cases, including crisis timing, safety-clearance language, privacy, diagnosis, and treatment boundaries. One synthetic test caught a gap in a refusal rule; I kept that failure on record and repaired the rule instead of loosening the test.
Current status: independent development and testing using synthetic cases and non-client information only. PHI processing, production clinical inference, and real-client clinical use remain disabled.
Implementation uses local LLM tooling and standard local development components such as Ollama, SQLite/ChromaDB, Docker, Flask, and Python/PowerShell test harnesses where useful.
AI Lab methodology: This work runs in a local, self-built environment I use to test multiple AI models under close human control. Actions remain bounded and require explicit approval, failed tests are preserved, and scope widens only after evidence from testing. It supports the methodology behind the clinical AI work; it is not a clinical product itself.
Toward a Breathing AI Lab is a manuscript in preparation for IEEE submission. It examines evidence-governed supervised autonomy in a local multi-model AI environment, including explicit human authority, bounded action, release gates, failure preservation, and recovery. It has not yet been published or accepted.
I’m interested in work where clinical expertise stays genuinely clinical, while informing safer products, better evaluation, and clearer decisions about risk, privacy, and human authority.
The clinical résumé and the independent AI project portfolio are kept separate on purpose: one documents the clinical career, the other the newer independent technical work.
Clinical experience, emergency psychiatric consultation, supervision, education, and a summary of the newer independent AI work.
One-page overview of guarded workflows, a self-authored risk taxonomy and evaluation cases, evidence sourcing, release gates, and failure-handling design.
For behavioral-health technology, clinical product safety, AI evaluation, clinical operations, risk review, and clinical-to-technical collaboration.