What changed for SLPs
- New audio uploads and text samples are routed to the ConductSpeech production analysis engine
- The app now checks both database and engine health before reporting the service as healthy
- Failed samples from the earlier outage remain visible as failed attempts, so clinicians can rerun them deliberately
- The updates page and marketing pages are fixed on mobile, so SLPs can read release notes from a phone or tablet
Clinicians should see fewer generic failed-analysis states when starting a new language sample.
If a previous sample failed with a fetch error, create a new analysis for that recording instead of relying on the old failed record.
Results still need transcript review and clinical judgment before they are used in documentation, eligibility discussions, or progress monitoring.
ConductSpeech now sends production transcription and language sample analysis work to a dedicated ConductSpeech engine service. For SLPs, the practical point is simple: new audio uploads and typed language samples should move through the analysis workflow instead of failing because the app cannot reach the analysis engine.
The app health check now verifies both the database and the engine, so service issues are easier to catch before they turn into confusing failed samples. This is especially important when a clinician is between sessions, reviewing a recording after school, or trying to finish documentation before a meeting.
If you uploaded a sample before this fix and saw a failed analysis, start a new analysis for that recording. Older failed records remain in the history as failed attempts so they are not silently changed, but new analyses are routed through the dedicated engine.
Clinical note
ConductSpeech is a clinical support tool, not a replacement for professional judgment. Review the transcript, speaker labels, and computed metrics before using results in a report, IEP discussion, or research dataset.
What we are fixing next
- Clearer sample status labels during upload, transcription, and analysis
- Better recovery guidance when an audio file cannot be processed
- More clinician-facing notes on how MLU, TTR, PGU, NDW, and NTW are calculated