The future of ASKLEMER
Turn clinical experience into reviewable intelligence at the point of care.
ASKLEMER’s long-term direction is a domain-specialized clinical platform for licensed practitioners: structured by longitudinal acupuncture SOAP records, designed to run inside the clinic, and built to preserve the clinician’s final authority over every suggestion.
Current status: this page describes a product roadmap, not generally available capabilities. Data rights, institutional review, record-level archive verification, de-identification design, model evaluation, and pilot deployment remain development milestones.
Why now
Use is rising while the professional base is contracting.
These figures describe the field; they do not predict ASKLEMER’s clinical performance, adoption, or business results. Sources appear below.
Planned clinical data foundation
Longitudinal. Structured. Outcome-linked. Professionally supervised.
Longitudinal records
Repeat encounters across a long-running teaching clinic can preserve the course of care rather than a single snapshot.
SOAP structure
Each encounter separates Subjective, Objective, Assessment, and Plan so evidence, reasoning, and clinical decisions remain distinct.
Outcome-linked
Progress across repeat visits can connect a practitioner’s earlier decisions to what was documented afterward.
Professionally supervised
Teaching-clinic records were created under licensed faculty supervision and consistent documentation standards.
Data status must remain explicit. The source institution holds the underlying records and patient relationship. ASKLEMER’s development plan requires definitive data rights, appropriate institutional review, and independent de-identification before the modeling team receives authorized de-identified material. The archive is believed to span about 45 years, pending record-level verification.
Product blueprint
Not a general chatbot—a complete clinical workflow.
Domain-trained
Developed around pattern differentiation, treatment principle, point selection, and longitudinal treatment records rather than open-web text alone.
Local-inference target
Designed to run on dedicated hardware inside the clinic’s chosen boundary so routine patient context and audit records can remain on site.
EHR-integrated
Ships as a standalone clinical workspace first, then connects to compatible EHRs through supported APIs, command-line tooling, or MCP—without replacing the existing system.
Practitioner-directed
Presents ranked candidate patterns and point suggestions with exposed rationale so the practitioner can independently review, edit, or reject them.
Auditable
Intended to record inputs, retrieved sources, generated rationale, and practitioner changes for internal quality review.
Multilingual
Begins with English and Chinese, with future languages and regional clinical material added only after appropriate authorization and evaluation.
At the point of care
Guidance inside the practitioner’s existing workflow.
01
Clinical record
Current encounter context arrives from the standalone workspace or an authorized EHR connection.
02
Protocol layer
Standard connectors retrieve only the information needed for the current task.
03
On-site inference
The domain model runs locally on dedicated hardware selected by the clinic.
04
Practitioner decision
The clinician reviews ranked options, rationale, and sources before deciding what to use.
This is a target architecture, not a HIPAA certification claim. HIPAA has no general appliance-certification program; compliance depends on the final deployment, contracts, policies, and use.
Phased roadmap
Build a defensible foundation before clinical deployment.
Governance, corpus, and model
Complete data rights and review, establish independent de-identification, normalize the archive, create training, validation, and held-out test sets, and run blinded evaluation with licensed practitioners.
Pilots and local deployment
Install local inference units at approved pilot clinics, connect compatible EHR workflows, and test citation fidelity, review burden, reliability, and workflow fit.
Broader clinical network
After validation, distribute updates to more clinical sites and explore settings that place conventional and integrative services together—then extend to massage therapy, moxibustion, and adjacent disciplines with separate data and safeguards.
Future access
One clinical intelligence layer, several ways to use it.
- Standalone workspace: SOAP drafting, cited retrieval, and practitioner review through a dedicated ASKLEMER website.
- Inside the EHR: practitioner-approved support in compatible systems such as Epic or athenahealth when vendor access and permissions allow.
- Clinic-local node: validated model updates running on hardware chosen by the institution.
- Education and standardization: tools for schools and teaching clinics to inspect the connection among findings, pattern, treatment principle, and point selection.
Public evidence
Field and governance references
- NIH NCCIH: acupuncture use and background
- Journal of Integrative Medicine: U.S. licensed-acupuncturist and education distribution
- U.S. HHS: HIPAA de-identification methods
- CMS: Medicare national coverage determination for acupuncture for chronic lower back pain
- U.S. Department of Veterans Affairs: acupuncture in Whole Health
- U.S. FDA: Clinical Decision Support Software guidance
External references provide field and regulatory context. They do not establish ASKLEMER’s safety, effectiveness, compliance status, or current availability.
Shape the future workflow
Bring real clinical practice into product development.
We are listening to licensed practitioners, clinics, and teaching programs.