Group sessions
One recording, many patients. Notes have to capture the session and each participant's part in it without mixing one person's information into another's record.
AI for Addiction Treatment Centers
AI for Behavioral Health & Addiction Treatment
We add AI to the software your program already uses: scribes for individual and group sessions, intake and assessment documentation, and treatment plan drafts that a clinician reviews and approves. Built for residential, outpatient and substance use disorder (SUD) programs.
- Documentation time per patient in our SureScribe build
- 60+ → ~20 min
- Documentation time per patient in our SureScribe build
- Nothing reaches the record until a clinician approves it
- Clinician-approved
- Nothing reaches the record until a clinician approves it
- Athenahealth and CharmHealth write-back shipped
- 2 EHRs
- Athenahealth and CharmHealth write-back shipped
- AI added to the software you already use
- Keep your tools
- AI added to the software you already use
The Opportunity
Less documentation, without replacing your EHR
Counselors and clinicians in addiction treatment spend a large part of every day documenting: intakes, biopsychosocial assessments, individual and group notes, treatment plans and reviews. Most AI documentation products for behavioral health come bundled with a full EHR, which means switching systems to get them.
Torch Solutions takes the other route. We add AI to the application, EHR and communication tools you already run, so your workflows, forms and reporting stay as they are. The AI listens, drafts and suggests; your clinicians review, edit and approve before anything is saved to the record.
This is the same engineering we shipped in SureScribe, an AI clinical documentation platform with clinician approval and EHR write-back that reduced documentation from more than an hour to about 20 minutes per patient. Behavioral health adds its own requirements, especially around SUD record confidentiality, and we design for those from the first conversation.
Business Challenges
What makes behavioral health documentation different
Long, structured assessments
Intakes and assessments have many fields and take a long time to complete. Clinicians need help filling them in, not a tool that decides for them.
SUD confidentiality rules
Substance use disorder treatment records are protected by 42 CFR Part 2 as well as HIPAA, with specific patient-consent requirements for disclosure.
Systems you can't replace
Programs already depend on an EHR, custom forms and patient-communication tools. A new platform is rarely an option; integration is.
Features & Capabilities
What we build for treatment programs
AI scribe for individual sessions
Record or dictate a session and receive a structured draft note in your program's format for the clinician to review, edit and sign.
Group session documentation
A session summary plus per-participant draft notes, with each note checked by the facilitator before it is filed to that patient's record.
Intake and assessment documentation
AI listens to the interview, suggests values for assessment fields and shows the supporting transcript evidence, so clinicians confirm rather than type.
Treatment plan drafting
Draft problems, goals and interventions from assessment findings and progress notes. The clinician approves every plan before it is used.
Integration with your EHR and tools
We connect to your existing EHR or application through its APIs and keep tools such as Emitrr in your workflow. We assess each system's integration options during discovery.
Consent-aware data handling
Access controls, audit logs and data flows designed around HIPAA and 42 CFR Part 2 consent, including which vendors may receive which data.
Business Benefits
Business value designed into the system
More time with patients
Drafts replace blank pages, so clinicians spend their time reviewing and correcting instead of typing.
No platform switch
Your EHR, forms, reports and patient-communication tools stay. AI is added where the work already happens.
Clinicians stay in control
Every note, field and plan is a draft until a licensed clinician approves it. The record remains theirs.
Built around your formats
Note types, assessment forms and treatment plans follow your program's templates and payer requirements.
How We Work
From first call to a pilot with your clinicians
Discovery
We map your note types, assessments, group formats, current systems and consent process, and identify where AI saves the most time.
Compliance and data-flow design
We define what is recorded, where it is processed and stored, who can see it, and how HIPAA and 42 CFR Part 2 consent apply, with your compliance lead.
Proof of concept
A working prototype on one workflow, such as individual session notes, evaluated with your clinicians on representative sessions.
Integration
We add capture, review and approval to your existing app or EHR and write approved content back.
Pilot
A small group of clinicians uses it in real sessions while we measure time saved, edit rates and issues.
Rollout and support
We extend to more note types and teams, monitor quality and keep improving.
Technologies We Use
A production stack selected for your requirements
We select speech and language models by measured performance on your sessions and by your compliance requirements.
- AssemblyAI
- AWS Transcribe
- Whisper
- OpenAI
- Anthropic
- LangChain
- RAG
- Python
- FastAPI
- React
- React Native
- PostgreSQL
- AWS
- FHIR
Who It's For
Programs we build for
Residential treatment centers
High documentation volume across intake, daily notes, groups and treatment plan reviews.
Outpatient and intensive outpatient programs
Individual and group sessions with recurring progress notes and plan updates.
SUD and co-occurring disorder programs
Programs whose records fall under 42 CFR Part 2 and need consent-aware data handling.
Behavioral health software vendors
EHR and practice-management companies adding AI documentation to their own product.
Why Torch Solutions
Why treatment programs work with Torch Solutions
We have shipped an AI scribe
SureScribe covers recording, transcription, multi-stage notes, clinician review and EHR write-back, with measured time savings.
We add to what you have
Our default is integration with your current software, not a new platform to migrate to.
Clinician-in-the-loop by design
Review, correction and approval are core features, not afterthoughts.
Honest about limits
We tell you where AI should not make the call, and we confirm licensing before working with proprietary clinical criteria.
Related Case Studies
Clinical documentation AI we have built
SureScribe AI Clinical Documentation Platform
HIPAA-aware AI scribe: clinical speech recognition, multi-stage LLM notes, clinician approval and Athenahealth / CharmHealth integration. Documentation cut from 60+ to about 20 minutes per patient.
Read Case Study →Engagement & Timelines
How an engagement runs
Start small, prove it with your clinicians, then expand.
Discovery
About 1–2 weeks
Workflows, systems, consent process and the best first use case.
- Workflow and data-flow map
- Integration assessment
- Plan and estimate
Proof of concept
About 4 weeks
One workflow working end to end, evaluated with your team.
- Working prototype
- Evaluation with clinicians
- Production plan
Pilot and rollout
Phased
Integration into your software, pilot group, then wider rollout.
- Integrated capture and review
- Audit logging and access controls
- Monitoring and support
This page is general information, not legal advice. Your compliance and legal teams make the final determination on HIPAA and 42 CFR Part 2.
Related Services
Combine this capability with the application, cloud, data, integration, and product engineering required to operate it reliably.
Frequently Asked Questions
Questions about Behavioral Health & Addiction Treatment
Do we have to replace our EHR to use AI?
No. We add AI to your existing application or EHR through its APIs or interface, and keep your current forms, workflows and reporting. If a system has no usable integration option, we tell you during discovery and propose alternatives.
How do you handle 42 CFR Part 2?
42 CFR Part 2 is the federal regulation that protects the confidentiality of substance use disorder treatment records and sets consent requirements for disclosing them. We design data flows, vendor choices, access controls and audit logs around your consent process, together with your compliance lead. This is general information, not legal advice.
Can the AI document group therapy sessions?
Yes. The design produces a session summary and a separate draft note for each participant, and the facilitator reviews each note so that one patient's information is not filed in another's record.
Will AI decide a patient's level of care?
No. AI can extract evidence from the conversation and suggest values for assessment fields, but a licensed clinician reviews, approves and makes the determination. We work with the criteria and assessment tools your program is licensed to use.
Does it work with Emitrr and our other tools?
Our approach is to keep the tools you rely on in place. For each system, including Emitrr, we review the available APIs and integration options during discovery before committing to a design.
How long does it take to get started?
Discovery takes about one to two weeks and a proof of concept about four weeks. Integration and rollout are delivered in phases after that.
Running a treatment program and exploring AI documentation? Contact Torch Solutions.
CustomSoftware DevelopmentCompany
Ready to Solve the Right Software Problem?
Talk with an experienced software team about your goals, workflows, users, integrations, and technical risks before you commit to a roadmap, architecture, or development budget.