Product
Resources
Pricing

Changelog

What we have shipped.

Every update to Audora, in order. We document what changed, what improved, and what got fixed - so you always know exactly what you are working with.

v0.9
Launch

Audora goes live - full platform launch

After several months of building and internal testing, Audora is now open for early access. All core functionality is live: the bot joins Zoom and Google Meet calls, transcribes with speaker labeling, generates all six structured outputs, and pushes into Cursor, Lovable, Bolt, Linear, and Notion.

Microsoft Teams support launched simultaneously on Studio and Agency plans. Three subscription tiers are available - Solo, Studio, and Agency - with a 14-day free trial on Studio for all new accounts.

NewZoom bot - joins via meeting link, real-time transcription, speaker labeling
NewGoogle Meet bot - joins via calendar event or direct link, full transcription
NewMicrosoft Teams bot - Studio and Agency plans, joins as external participant
NewSix outputs per call - transcript, client brief, requirements doc, PRD, task breakdown, SOW
NewCursor push - PRD written to project repo as structured spec file
NewLovable push - pre-loaded project link generated from PRD
NewBolt push - pre-loaded project link formatted for Bolt's prompt structure
NewLinear push - creates project with issues from task breakdown (Studio and Agency)
NewNotion push - creates structured page with complete project package (Studio and Agency)
NewThree subscription plans - Solo ($49/mo), Studio ($99/mo), Agency ($199/mo)
New14-day free trial - Studio plan, no credit card required
v0.8
Improvement

Output quality overhaul and processing speed

After extensive testing with early beta users, we rebuilt the core output generation pipeline. The biggest change: three separate, focused AI calls instead of one large prompt - one for extraction, one for structuring, one for the client-facing SOW. The result is significantly more accurate outputs and much cleaner formatting across all six document types.

Processing time also dropped from an average of 3.5 minutes to under 60 seconds for a typical one-hour call. This was the single most-requested improvement from beta users.

ImprovedProcessing speed - average time from call end to full output reduced to under 60 seconds
ImprovedPRD quality - requirements extraction rebuilt with dedicated structured prompt
ImprovedSOW formatting - cleaner section structure, better scope language, less generic phrasing
ImprovedTask breakdown - time estimates are now better calibrated to the actual complexity discussed in the call
ImprovedSpeaker labeling accuracy - reduced misattribution in calls with more than three speakers
v0.7
Improvement

Linear and Notion integrations, team seats

Studio and Agency plan users now have access to Linear and Notion push integrations. The Linear integration creates a full project with issues from the task breakdown. The Notion integration creates a structured page with the complete output package - PRD, requirements, checklist, and SOW embedded as blocks.

Team seat management is also live on this release. Account owners on Studio and Agency plans can now invite team members, set permissions, and manage access through the account settings.

NewLinear integration - project and issue creation from call outputs (Studio and Agency)
NewNotion integration - structured project page with all outputs embedded (Studio and Agency)
NewTeam seat management - invite, remove, and manage team members from account settings
ImprovedDashboard layout - call history, outputs, and integrations reorganized for clarity
ImprovedBolt push - prompt structure refined based on Bolt's updated input format
v0.6
Fix

Bot reliability fixes and edge case handling

A focused release addressing the most common failure modes reported during early beta. The bot now handles waiting rooms, late starts, and host-muted participants correctly. Several edge cases in the transcription pipeline that caused partial or missing outputs on calls over 90 minutes were also resolved.

FixedWaiting room handling - bot now waits correctly and joins when admitted, rather than timing out
FixedLong call transcription - fixed partial output generation on calls longer than 90 minutes
FixedSilent participant handling - calls where some participants never spoke no longer cause output failures
FixedCursor push path - fixed edge case where spec file was written to wrong directory in monorepos
FixedSession timeout - authentication tokens were expiring mid-call on long sessions; resolved
ImprovedError notifications - processing failures now send an email with the specific reason rather than a generic error
v0.5
Beta

Private beta - first external users

Audora opened to the first cohort of external beta users. Core functionality at this stage: Zoom and Google Meet bot, full transcription, and the PRD and task breakdown outputs. Cursor push was the only integration live at this point. Feedback from this cohort directly shaped v0.7 and v0.8.

BetaFirst 50 external users - private invite-only beta begins
BetaZoom and Meet bots - core transcription and speaker labeling functional
BetaPRD and task breakdown - first two output types available
BetaCursor integration - spec file push to project repo

See our roadmap