FAQ
Frequently asked questions about Prism conversation analysis — supported file formats, privacy, data security, and how AI analyzes team communications.
What should my transcript look like?
Plain text (.txt) is the simplest option. Save or upload a .txt file where each line starts with the speaker name, a colon, and what they said (e.g. "Host: Welcome…"). No special formatting is needed. WhatsApp exports, VTT/SRT, CSV, and JSON use the same idea — Prism detects the wrapper automatically. If a line has no speaker label, it may not be attributed correctly.
How do I get a transcript from Microsoft Teams?
Microsoft Teams lets you download a transcript of your meeting, which includes who said what. You can open that file, copy all the text (including speaker names), paste it into a text editor or notes app, save it as a text file, and upload it into Prism for analyzing it.
What file formats are supported?
Prism supports WhatsApp (.txt) exports, Zoom/Teams VTT/SRT subtitles, CSV with speaker and text columns, and JSON AI transcripts (e.g. Gemini, Whisper). Formats are detected automatically. Each line or message needs a speaker label and message text. Prism does not fix transcription errors, merge or split formats, or support arbitrary custom formats.
Can I analyze conversations in languages other than English?
Yes. Prism supports transcripts in any language — no need to translate before upload. Upload plain text, WhatsApp exports, VTT/SRT, CSV, or JSON as recorded. Analysis results (topics, insights, participation metrics, reports) are delivered in English. Best results with one primary language per transcript; bilingual conversations are supported; heavy mixing of three or more writing systems may reduce quality.
What if my VTT/SRT file is missing speaker names?
If your VTT or SRT file is missing speaker identification in more than 50% of messages, you can cancel and add speaker names manually for best accuracy, or continue with auto-generated labels (Speaker 1, Speaker 2, etc.). Prism needs speaker labels and message text; it does not fix transcription, merge/split formats, or support arbitrary custom formats.
Why does my analysis show only one speaker when there are many?
For VTT/SRT files, Prism detects speakers from labels in the file (for example VTT voice tags like v Speaker 0, or lines like "Speaker 1:"). If most lines have no speaker label, it may only recognise one speaker. For best results, edit the transcript so every line has a speaker label.
What file size limits apply?
We prioritize conversation length over file size. Pro users can analyze up to about 3 hours of conversation per upload; free users can analyze up to about 10 minutes. These limits are approximate and can vary slightly by file format (TXT, CSV, WhatsApp, JSON, etc.). Please contact us if you have transcripts longer than these limits.
What does the analysis show?
Prism analyzes conversations for an executive summary with efficiency scores, topic flow / timeline (narrative arc and key moments), participation analysis, standout contributors, sentiment analysis, topic contribution by speaker, decisions and actions, and blockers and risks.
Who can benefit from Prism transcript analysis?
Prism is used by team leaders, project managers, researchers, communication analysts, HR professionals, coaches, consultants, educators, and organizations worldwide. It's valuable for anyone who wants to improve meeting effectiveness, understand communication patterns, analyze team dynamics, evaluate training sessions, review client calls, or make data-driven decisions about collaboration and engagement. The Topic Flow Analysis feature provides intelligent insights that reveal hidden dynamics and conversation evolution—going far beyond simple meeting summaries to show participants, facilitators, and leadership not just what was discussed, but how the conversation evolved, where critical shifts occurred, and what patterns emerged.
How is my transcript data used?
Your transcript data is processed securely and used solely for generating your analysis. We do not store your raw transcript content at all. Analysis results are saved to your account optionally so you can access them later, but you can delete them anytime from the My Reports page.
Is my data private?
Yes. Your raw transcript is never saved - we only process it to generate your analysis. The analysis results can only be saved if you choose to save them. Once saved, you can share your analysis as you wish. When you share a report, only the analysis results are shared - not the raw conversation data. Shared reports are protected from search engines and AI bots, cannot be embedded in iframes, and include bot detection to prevent automated scraping. You can also restrict the number of views (10, 50, or 100 views) when creating share links. Share link expiration: Free (24h, cannot extend), One-Time (1 month from first analysis, absolute expiry), Pro Monthly (up to 30d, can recreate to extend), Pro Annual (up to 1 year, can recreate to extend).
Can I delete my account?
Yes. You can permanently delete your account and all associated data at any time from your account settings. This action is immediate and irreversible.
Can I try it for free?
Yes! Sign up for a free account to start analyzing your conversations. No credit card required.