Aria

From document to finished audio, in your own voice.

Turning your content into audio shouldn’t mean booking a studio

You already have the material: the report, the dossier, the knowledge article, the campaign brief. Turning it into a polished narration, a two-host podcast, or an executive briefing usually means a script, a booth, a producer, and a few days you do not have. Aria is the production studio around the voice engines: it writes the script, casts the speakers, synthesizes each line, mixes in the background bed, and manages the voices themselves, from off-the-shelf synthesized voices to a leader’s own cloned voice trained from their recordings. It runs the visual counterpart through Soundstage too: infographics, social media sets, cover art. Voice is a brand asset, not a one-shot transaction.

How it differs

A raw voice engine is text in, audio file out. Aria is the studio around them. Voice profiles as first-class records (consent, language, accent, tier, persona bio). Every clip versioned. Voice tracks mixed with background beds and ducking. Scripted multi-voice productions where each speaker has a separate voice and personality.

A single-track audio editor expects a producer at a desk. Aria is built for content that is generated, not edited. Start with source material, pick a format, and Aria writes the script, casts the speakers, synthesizes each line, and mixes the result. Every speaker line is independently re-renderable.

Who it’s for

Marketing teams generating weekly podcast-style episodes from blog posts, market analyses, or campaign briefs without booking a studio.

Internal communications and L&D turning SOPs, knowledge bytes, and reference guides into training modules.

Podcast producers and content creators converting long-form research, dossiers, or industry reports into ten-minute conversational episodes.

Sales and customer success generating short executive briefings from Forge campaign research before a customer call.

Voice owners (executives, hosts, instructors) who want their own cloned voice through Voice Trainer’s teleprompter recording, take assembly, and Instant or Professional cloning workflow.

What it does

Soundboard runs in two stages, on purpose. Stage one is script generation: source material flows in, the platform writes a multi-voice script as an array of speaker, text, and stage direction, and the user reviews and edits section-by-section. Stage two is audio synthesis: only after the script is approved do the casted voices render each speaker’s lines and the result mixes with the bed of your choice. Splitting script from synthesis means a bad fact in the script costs seconds to fix, not a dollar in TTS credits.

Voice profiles are configured voice identities with provider, speed, pitch, style, language, and accent. Pick “Aria Host” instead of “OpenAI Nova at 1.0x with formal style” and get the same voice every time. Cloned voices live in the same model with requires_consent set true and an approval workflow that fires on every downstream use.

The voice library is versioned. Every synthesis records characters used, estimated cost, voice profile snapshot, storage entry, and duration. Re-synthesize with overrides and the new clip links back to the original as a variant. The cost ledger is real.

Soundstage is the visual counterpart. Same project structure, same source-driven workflow. Infographics, social media sets, cover art. Brand kits flow through visual production the same way they flow through voice.

How it fits the ecosystem

Aria reads from anywhere on the platform. A Foundry document narrated as a briefing. A Forge dossier voiced as a five-minute analyst note. An OS knowledge byte turned into a training module. An Orbit account review summarized for the sales rep on the way to the call. The brand kit attached to the customer in Orbit drives the voice profile, the script tone, and the templating.

Where this product is in development

Aria is feature-complete and in final validation against customer workloads. What is live today is in active use; production deployment timing and the next capabilities are firming up with the beta cohort.

[Join the Beta Cohort]

Beta products are feature-complete and in final validation against customer workloads. Early access available; production deployment timing is on the roadmap.