Segriva Sakura Studio
Help & Documentation
Everything you need to know about using Sakura Studio — from ingesting your first video to publishing across every platform. Use the section links below to jump to any feature.
Dashboard
The Dashboard gives you a real-time bird's-eye view of your workspace. It shows active render jobs, recently published content, queue depth, AI token consumption, and storage utilisation. All numbers are scoped to your active community/tenant context — switching communities instantly refilters every metric.
- Live render queue depth with estimated completion time
- Recent publish history with platform icons
- AI token and storage usage gauges
- Quick-action cards to jump into any module
Content Feed
Content Feed is the aggregation hub. Connect RSS feeds, YouTube channels, social accounts, and webhook sources. Every incoming item is parsed, deduplicated, and indexed. From here you can one-click convert any item into a Studio job — triggering transcription, translation, and AI article generation in seconds.
- Multi-source RSS + YouTube + social aggregation (yt-dlp for YouTube — RSS deprecated by YouTube in 2024)
- YouTube stats: views, likes, duration, estimated watch hours, subscriber count
- Watch-hours formula: views × duration(s) × 50% avg completion ÷ 3600 — estimate only; real data requires YouTube Studio
- Twitter/X and Facebook require OAuth — connect via Settings → Integrations
- Automatic deduplication and priority scoring
- One-click "Create Studio job" from any feed item
- Filter by source, language, topic, and date range
Content Studio
Content Studio is where AI meets your editorial workflow. It reads the transcribed audio/video, generates multi-language articles, summaries, social captions, and SEO metadata — all from a single video upload. Every output is editable before publishing.
- AI article generation in 20+ languages
- Auto-SEO titles, meta descriptions, Open Graph tags
- Social captions sized for X/Twitter, Instagram, LinkedIn
- Translation pipeline powered by your configured LLM
- Rich text editor with version history
AI Video Studio
AI Video Studio automates the most tedious parts of video production. Upload a raw recording and the system transcribes it, identifies chapters, adds burned-in captions, applies your brand overlay, and queues it for rendering — all without manual effort.
- Whisper-powered transcription (local or cloud)
- Auto-chapter detection and scene markers
- Burned-in multilingual subtitles
- Brand kit overlays (logo, lower-thirds, end-cards)
- Timeline-based clip segment extraction
Media Library
Media Library is the central media repository. All uploaded originals, proxy files, rendered outputs, and AI-generated thumbnails live here. Files are organised by source, content type, and date. You can preview video proxies, download originals, and drag assets directly into jobs.
- Storage-backed file browser with proxy previews
- Filter by type (video, audio, image, document)
- Drag-to-job: attach any asset to an active Studio job
- Auto-thumbnail generation on upload
- Storage usage breakdown per asset type
Bundle Render
Bundle Render lets you batch multiple video segments, overlays, and captions into a single composition and export it as a platform-ready file. Define presets (16:9 for YouTube, 9:16 for Reels/Shorts, 1:1 for feed) and process them all in one job.
- Multi-format presets: 16:9, 9:16, 1:1, 4:5
- Audio normalisation and loudness targeting
- Caption burn-in with font and position control
- Watermark and logo overlay with opacity control
- Output to MP4/H.264 or H.265 with configurable bitrate
Export Queue
Export Queue gives you a transparent view into every active, pending, and completed processing job. Each job shows live progress, logs, and estimated completion. Failed jobs surface the error and offer a one-click retry. You can cancel or reprioritise any job in the queue.
- Live job progress with percentage and ETA
- Streaming log output for debugging
- One-click retry on failed jobs
- Job cancellation and priority reorder
- Filter by status: pending, running, done, failed
Live Streams
Live Streams handles real-time RTMP/HLS streams. Feed in a live broadcast URL and Segriva captures it, segments it into clips, transcribes speech in near-real-time, and buffers segments for downstream processing. Ideal for live news, sports commentary, and event coverage.
- RTMP and HLS stream ingestion
- Near-real-time speech transcription via Whisper
- Auto-segmentation at scene cuts and silence
- Buffered clip library per live session
- One-click "Create article from segment"
Publishing
Publishing connects your finished content to every platform in your distribution graph. Select the platforms, choose the content variant (video, article, social caption), set a publish time, and hit Publish. Each platform's status is tracked individually so a failure on one never blocks the rest.
- YouTube, Facebook, Instagram, TikTok, X, LinkedIn support
- Scheduled publishing with timezone awareness
- Per-platform status tracking (pending, live, failed)
- Automatic retry on transient API errors
- Publish history with platform-level analytics link
AI Providers
AI Providers is the nerve centre for all AI capability configuration. Split into three sections: Content Generation (LLMs for article writing, translation, summarisation), Media Processing (Whisper / STT for transcription), and AI Video Studio (text-to-video, image generation). Each provider can be independently configured with API keys and model selection.
- Content Generation: OpenAI, Anthropic, Google Gemini, Groq, local Ollama
- Media Processing: OpenAI Whisper (cloud), local Whisper sidecar
- AI Video: RunwayML, Kling, Wan, Stability AI
- Per-category API key vaulting — never logged
- Model selection and temperature/token controls per provider
Integrations
Integrations manages your OAuth connections to social and video platforms. Connect YouTube, Facebook, Instagram, TikTok, X (Twitter), and LinkedIn. Each platform shows the connected account name, permission scope, and connection health. Revoking a connection removes all stored tokens immediately.
- OAuth 2.0 secure connection flow per platform
- Token refresh handled automatically in the background
- Permission scope display (read, write, manage)
- Instant revoke — clears tokens from secure storage
- Multi-account support per platform (coming soon)
Settings
Settings houses workspace-wide configuration: local Whisper sidecar status, community context files, storage paths (desktop mode), and app information including the current version, build number, and release date. The desktop edition also surfaces the Python install status with a one-click link to get started.
- Local Whisper sidecar install and status indicator
- Community context files for AI persona and tone
- Storage path configuration (desktop mode)
- App version, build number, and release date display
- Python environment status with install guidance
Still need help?
Reach out to the Segriva team and we'll get back to you as soon as possible.
Contact support