Sam Liang is Otter's co-founder and chief executive, identified in its company announcements. Otter's work spans transcription and meeting assistance. His profile explores how spoken conversation becomes a searchable record, and why recognising words, attributing speakers and extracting an agreed action are separate technical and editorial tasks.
- Person
- Sam Liang
- Company
- Otter.ai
- Area of work
- Transcription and meeting assistance
The product direction
Otter’s own company announcements identify Sam Liang as co-founder and CEO and describe a product built around transcription and AI meeting assistance. This profile considers that product category rather than repeating company growth claims as an independent assessment of success.
Public record: Otter.ai: founder biography or company record
What can go wrong between speech and a summary?
An audio system may mishear a name, assign a sentence to the wrong speaker or miss a negation. A later summarisation step can amplify that error by presenting it as a concise conclusion. An apparently polished note therefore needs to be checked against the underlying conversation when accuracy matters. Otter's company announcements describe its meeting-assistance direction; the relevant evaluation should separate recognition quality from the usefulness and faithfulness of the summary built on top of it.
Why live and after-meeting uses differ
During a meeting, a transcript may help a participant follow the discussion. Afterwards, a team may use the record to answer questions or update another system. Those workflows have different tolerances for delay and error. Our analysis is that buyers should define the intended use before comparing tools. A fast live transcript is not automatically the best final decision record, while a carefully revised summary may be too late for someone who needs immediate assistance.
How to run a fair transcription comparison
Use the same authorised recording across candidates, with a mix of clear speech, interruptions and relevant terminology. Prepare a human-checked reference for a short section and separately identify the meeting's actual decisions. Compare errors that change meaning, not just spelling differences. Review export, access and correction workflows as well as the output. This is a proposed comparison method, not a report of Otter's performance. It makes the test reproducible and keeps company-reported growth figures separate from evidence about your own meeting needs.
Questions about Sam Liang
Is transcription accuracy the same as summary accuracy?
No. A correct transcript can still be summarised misleadingly, and a transcript error can be repeated in the summary. Both stages need attention.
Should a meeting assistant be tested on a polished demo recording only?
No. Include the accents, terminology, interruptions and audio conditions relevant to the intended meetings, using material you are authorised to process.
Follow the evidence.
This AI-assisted profile combines the linked public record with Promptwire's editorial analysis. It is not an interview or a hands-on product review. Suggested tests are proposals, not reported results. Dated roles describe the period identified in the text. Company claims are attributed to their source.
Public sources support the factual background. Analysis and illustrative exercises are identified in the text. How we assess AI · Editorial standards
Spot a factual error or an outdated detail?



