sales-happyscribe（销售 快乐抄写员）

HappyScribe Platform Help

Step 1 — Gather context

If references/learnings.md exists, read it first for accumulated platform knowledge.

1. What's your primary goal?

- A) Transcribe audio/video files (batch transcription) - B) Generate subtitles/captions for video - C) Translate transcripts or subtitles into other languages - D) Order human transcription for high-accuracy needs - E) Integrate HappyScribe API into a pipeline - F) Compare HappyScribe to alternatives for my use case - G) Troubleshoot a specific issue (accuracy, editor, exports)

1. What type of content are you transcribing?

- A) Meetings / calls - B) Interviews / podcasts - C) Legal / compliance recordings - D) Media production (film, TV, marketing videos) - E) Research / academic - F) Mixed / high volume

1. What plan are you on (or considering)?

- A) Free trial (10 min) - B) Basic ($17/mo — 120 AI minutes) - C) Pro ($29/mo — 600 AI minutes) - D) Business ($89/mo — 6,000 AI minutes) - E) Enterprise (custom) - F) Not sure yet

Skip-ahead rule: if the user's prompt already contains enough context, skip to Step 2.

Step 2 — Route or answer directly

When routing, provide the exact command: "This is a {problem domain} question — run: /sales-{skill} {user's original question}"

Otherwise, answer directly below.

Step 3 — HappyScribe platform reference

Read references/platform-guide.md for the full platform reference — modules, pricing, integrations, data model, workflows.

If the question involves the API, also read references/happyscribe-api-reference.md.

Answer the user's question using only the relevant section. Don't dump the full reference.

Step 4 — Actionable guidance

You no longer need the platform guide — focus on the user's specific situation.

Plan selection framework:

• One-off or light use (<2 hrs/mo) → Basic ($17/mo for 120 min, or $8.50/mo annual)
• Regular transcription (up to 10 hrs/mo) → Pro ($29/mo for 600 min, 3 seats included)
• High-volume production (up to 100 hrs/mo) → Business ($89/mo for 6,000 min, 5 seats, style guides)
• Overflow credits: $0.20/min ($12/hr) once monthly allotment is exhausted
• Human transcription: $2.00/min ($120/hr) — order separately from subscription

AI vs Human decision:

• AI (85% accuracy) → drafts, internal use, content with clear audio and common accents
• Human (99% accuracy) → legal, broadcast, publication, heavy accents, poor audio quality

Accuracy optimization:

1. Upload highest quality audio (minimize background noise, crosstalk)
2. Use custom glossaries (Business+) for domain-specific terms
3. Use style guides for formatting consistency across team projects
4. Review speaker labels after transcription — diarization is imperfect

If you discover a gotcha, workaround, or tip not covered in references/learnings.md, append it there.

Gotchas

*Best-effort from research — review these, especially items about plan-gated features and integration gotchas that may be outdated.*

• No live/real-time transcription. HappyScribe is upload-only — it does not join meetings or capture audio live. For live transcription, use Fathom, Fireflies, or Otter.
• 85% AI accuracy requires manual cleanup. Expect significant corrections with background noise, multiple speakers, or strong accents. Budget editor time into your workflow.
• Free trial is 10 minutes total, not recurring. Exports are watermarked on free tier. You cannot evaluate properly without paying.
• Pricing can change suddenly. Multiple users report unexpected price increases. Lock in annual billing to protect your rate.
• Style guides are Business-only. Teams needing formatting consistency across projects must be on Business ($89/mo) or Enterprise.
• API rate limit is 200 req/hour. Lower than most transcription APIs — queue requests and cache responses for high-volume pipelines.
• Editor can lose unsaved work. Some users report page refresh bugs that wipe edits. Save frequently and export drafts.
• Human transcription is billed separately. The $2.00/min cost is on top of your subscription fee, not instead of it.

Related skills

• /sales-note-taker — Comparing AI meeting note-takers and conversation intelligence platforms (Fathom, Fireflies, Gong, etc.). HappyScribe is a batch transcription option, not a live meeting tool.
• /sales-sonix — Sonix platform help (closest competitor — batch AI transcription, 53+ languages, REST API, SOC 2/HIPAA)
• /sales-integration — General tool integration patterns (Zapier, webhooks, iPaaS)
• /sales-do — Not sure which skill to use? The router matches any sales objective to the right skill. Install: npx skills add sales-skills/sales --skill sales-do

Examples

Example 1: Choosing HappyScribe vs Sonix for subtitle production

User says: "I produce marketing videos in 5 languages and need subtitles. Should I use HappyScribe or Sonix?" Skill does:

1. Notes HappyScribe supports 120+ languages vs Sonix's 53+ — wider coverage for multilingual workflows
2. Compares subtitle editors: HappyScribe has styling/positioning controls and more export formats (SRT, VTT, STL, Premiere XML, FCP XML); Sonix has video burn-in
3. Compares pricing: HappyScribe Pro $29/mo for 600 min vs Sonix Premium $22/mo + $5/hr — HappyScribe is cheaper at moderate volumes
4. Recommends HappyScribe for multi-language subtitle workflows with extensive format needs

Result: Clear recommendation based on language coverage and subtitle feature comparison.

Example 2: Building an API pipeline for transcription

User says: "I want to auto-transcribe uploaded recordings using the HappyScribe API and export as SRT." Skill does:

1. Reads the API reference for endpoint details
2. Sketches the pipeline: Upload via signed S3 URL → Create Order with is_subtitle: true → Poll or webhook for completion → Create Export with format: srt → Download from export URL
3. Notes: use Orders endpoint (Transcriptions endpoint is deprecated), 200 req/hr rate limit, Bearer token auth
4. Flags: confirm: true on order creation to auto-submit, or create then confirm separately for review

Result: Working pipeline architecture with HappyScribe-specific endpoint details.

Example 3: Transcription accuracy is terrible with accented speakers

User says: "HappyScribe keeps getting names and technical terms wrong — the transcript is unusable for our legal team." Skill does:

1. Explains 85% AI accuracy is typical — for legal-grade work, recommends human transcription ($2.00/min, 99% accuracy)
2. For AI transcription improvement: create a custom glossary with names and technical terms (Business+ required), upload highest quality audio, ensure clear speaker separation
3. Suggests a hybrid workflow: AI transcription first for speed, then order human proofreading on critical sections

Result: Practical fix with cost/accuracy trade-off guidance.

Troubleshooting

AI transcription accuracy drops with multiple speakers or accents

Symptom: Transcript has frequent word errors, especially with non-native English speakers or cross-talk Cause: HappyScribe's AI engine (Whisper-based) struggles with overlapping speech, strong accents, and domain-specific terminology Solution: Upload the highest quality audio available. Create a custom glossary (Business+) with proper nouns, technical terms, and acronyms. For critical content, use human transcription ($2.00/min). If accuracy is consistently below 70%, the audio quality may be the limiting factor — consider using a dedicated recording tool with noise cancellation.

Editor loses changes on page refresh

Symptom: Edits disappear after the browser refreshes or the page reloads unexpectedly Cause: Known intermittent bug where the editor fails to auto-save before a page refresh Solution: Save manually and frequently (Ctrl/Cmd+S). Export a draft after major edits as a backup. Use Chrome or Firefox (most tested). Avoid editing on unstable internet connections. If the issue persists, contact HappyScribe support with your browser version.

Translation quality is poor or nonsensical

Symptom: Translated transcripts contain gibberish, made-up content, or completely wrong translations Cause: HappyScribe changed their AI translation model and quality regressed for some language pairs Solution: Review translations immediately after generation — don't publish without human review. For critical translations, order human translation instead of AI. If a specific language pair consistently fails, report it to HappyScribe support and use a dedicated translation service in the interim.