Account-tied work
Saved sessions, study sets, and exports are organized around signed-in user workflows.
Privacy and data handling
Clear controls for speech samples, transcripts, and clinical language data.
ConductSpeech is built for SLPs, districts, clinics, and research teams that need modern language sample analysis without losing control of sensitive data.
Last updated: May 24, 2026.
What ConductSpeech processes
The product works with the material needed to analyze a language sample. Customers decide what identifying details to include.
Audio recordings
Audio is uploaded or recorded by the clinician and processed to produce a transcript through the ConductSpeech workflow.
Transcripts
Transcripts may be generated from audio or entered manually. Clinicians can edit transcripts before final analysis.
Language metrics
ConductSpeech calculates measures such as MLU, NDW, TTR, PGU, narrative scores, and protocol-specific scores.
Reports and exports
Draft reports, CSV exports, SALT-compatible files, and research exports are generated from the customer's own samples and settings.
Protected workflow
A plain path from sample to saved work
ConductSpeech keeps the privacy story understandable: add only the sample you are allowed to analyze, review the output, save what is useful, and delete work that is no longer needed.
1
Add an authorized sample
Use a recording, transcript, or saved session that your team is allowed to analyze.
2
Review before relying on results
Confirm speaker turns, identifiers, and transcript edits before using the analysis.
3
Use controlled outputs
Reports, study sets, and exports stay tied to your ConductSpeech account workflow.
4
Keep only what you need
Use participant IDs where possible and delete saved work when it is no longer needed.
Privacy-first workflow choices
ConductSpeech gives teams room to follow minimum-necessary data practices without making the product harder for clinicians to use.
- +Use participant IDs or caseload IDs when possible.
- +Avoid putting names, dates of birth, or addresses directly into transcript text unless required by your workflow.
- +Delete saved sessions that are no longer needed.
- +Use team roles to limit who can manage members and shared caseloads.
- +Contact ConductSpeech for compliance review before using PHI in a regulated workflow.
Minimum necessary details
Teams can use participant, student, or clinic IDs instead of placing names in transcripts.
Reviewable outputs
Clinicians can inspect transcript edits, analysis results, and export readiness before sharing.
Deletion support
Saved sessions can be deleted by the user or handled through a support request.
Data retention
Saved sessions remain available in the customer account until they are deleted or the customer requests account deletion. Temporary processing files are kept only as long as needed to complete the workflow. Customers can request deletion by emailing info@conductscience.com.
Third-party services
HIPAA-ready workflows use protected ConductSpeech processing. Some non-HIPAA or customer-approved workflows may use additional processors for report support or advanced analysis only when configured for that use case. We do not sell customer data or use customer samples to train external models.
HIPAA and compliance
ConductSpeech is designed for HIPAA-regulated workflows, but every covered entity has its own compliance requirements. Organizations that plan to use ConductSpeech with PHI should complete a compliance review with us before production use.
Read the HIPAA overview and security controls pages for more detail.
Privacy resources
Compliance support
Need a privacy or security review?
We can walk your team through account access, sample handling, deletion support, and the agreement path before production use.