Product brief / In development
Creator Flow
A repeatable tool for independent creators: turn your own material into narrated social videos, with an approved script and a voice you choose.
The problem we're starting with
Creators often have photos, footage and a point of view before they have a video they are ready to share. They may want a different public narration voice, pronunciation or delivery, even when publishing in one language. Finding a narrator, coordinating retakes, timing captions and editing the story takes work every time.
The founder has described this problem directly. Its prevalence, severity and willingness to pay among other creators have not been validated. An accent is not a defect; the creator chooses how the finished story sounds.
A focused first audience
Independent travel and UGC creators who already own media, publish repeatedly and find spoken storytelling easier than editing. Small hospitality teams are a later adjacent audience, not evidence of an existing customer base.
The product hypothesis
Own photos and clips, a spoken account and an optional style reference become a reviewable script and edit. The creator selects a licensed narrator or a permitted own voice in their chosen language, checks the meaning and delivery, then approves the video. Optional translation comes after the first approved version. A faceless voiceover does not need lip-sync or a literal image for every sentence.
Start with a short Instagram Reel. A fuller Facebook edit can retain more of the author's commentary, and longer YouTube stories are a separate direction to test. These are intended export formats, not verified publishing integrations.
Why now, and why not just translate an existing video?
YouTube reported in September 2025 that creators who uploaded multi-language audio received over 25% of watch time, on average, from non-primary languages (data as of July 2025). This is platform-reported evidence of cross-language viewing among adopters, not a causal uplift, a Facebook benchmark, or proof that creators will buy our product.
Meta already offers Reel translation, dubbing and optional lip-sync. Translation alone is therefore not a defensible differentiation. We need to prove that starting from rough photos and a spoken story, choosing a voice and style, and reviewing reusable language versions saves meaningful work beyond native platform tools.
Why this is more than a voice tool
Speech generation alone does not choose the right memory, check a translation or respect a content brief. Our proposed differentiation is the continuity between the original story, script, media, captions and final review. That advantage is a hypothesis to test, not a claimed competitive moat.
Why we're evaluating ElevenLabs
Take one approved story about a coastal stay. The creator keeps the photos and meaning, but chooses a different licensed narrator for the final cut. That same process can narrate a campsite story, a product demonstration or a UGC brief. A second language is optional.
- Generate and compare: use text-to-speech for the same approved script in different voices and compare pronunciation, pace and tone.
- Revise and export: regenerate only changed passages, check caption timing and render a reviewed video. Test a second language only after the first version works.
- Measure: track generations and retakes, credits per approved export and time to approval across several stories. Use these results to test the economics of a subscription.
Grant credits would support this product validation, not provider-credit resale. Live API output and the complete creator workflow are not yet validated. This is an intended integration, not an awarded grant or a partnership.
Voice conversion may preserve an accent. Script-to-speech with a selected narrator is a separate path. We do not promise automatic accent removal or clone anyone's voice without explicit consent.
What exists, and what comes next
- Local prototype
- English transcription with whisper.cpp; prepared-script rendering using selected media; timed captions; portrait, feed and landscape versions; audio caching and stale-audio guards.
- Next validation
- One permitted real story from input to approved export, including script review and a different narration voice. Then test retakes, timing and the creator-facing workflow with a small pilot.
- Public sample
- A silent seven-second illustrative scene and three fictional example scripts. Where supported, browser playback uses a local device voice, not ElevenLabs. Neither is a customer result.
A business, not a one-off video
CONSENS PTY LTD is developing its own commercial software under the Personal AI Systems brand. The proposed model is a monthly subscription for the creation workflow, with a defined generation allowance and optional, explicitly approved extra usage. The customer buys a repeatable way to create and localize videos; it is not provider-credit resale. No subscription or charge is available on this website.
Each reviewed language script could generate narration through the ElevenLabs API. Our application would own the story, media selection, templates, approval, captions and export. Reusing approved audio would avoid unnecessary regeneration. We must measure the cost of all language versions and retries, plus transcription, translation, rendering, storage and support, against subscription revenue. The grant would fund evaluation, not conceal an unviable paid cost structure. Pricing, unit economics and willingness to pay remain to be tested.
Market sizing: a transparent planning scenario
Source baseline, not a current census: Goldman Sachs, 19 April 2023, estimated 50 million creators worldwide, with about 4% earning over US$100,000 a year. That implies approximately 2 million professional creators. It does not identify multilingual travel creators or demand for this software.
Revenue assumption: US$15 per paying creator per month, or US$180 annually, is a modelling input only. It is not a validated price, product offer, committed plan or measured average revenue.
- TAM ceiling scenario: US$360 million annually
- 2,000,000 professional creators × US$180. This assumes every creator in that broad cohort fits and pays; actual product-addressable demand could be much smaller. It is not a measured TAM or a valuation.
- SAM sensitivity: US$3.6–18 million annually
- If 1–5% of that cohort fits a repeat-publishing, voice-led travel/UGC workflow, that is 20,000–100,000 hypothetical subscribers at US$180 each. Those segment shares are unvalidated assumptions, not research findings or a reachable mailing list.
- Early adoption scenario: US$18,000–180,000 annually
- 100–1,000 paying creators × US$180. This is arithmetic, not existing traction, a forecast or an evidence-based SOM. A credible obtainable-market estimate needs measured acquisition, retention and paid conversion.
How we'll test the market
Proposed next test: recruit a pilot group of ten independent travel and UGC creators who already publish and want more control over narration. Compare one approved story narrated in a different voice against their current process. Test a second language separately when a participant actually needs it. Measure completion without developer help, hands-on time, factual and language corrections, approved-video cost and repeat use. Test willingness to pay separately; compliments or sign-ups are not purchases.
Neither the number of people who want this specific workflow nor a paid conversion rate is known today. Interviews, completed pilot tasks and paid retention would replace the assumptions above. No interviews, pilot customers or revenue are claimed here.
Engineering foundation
The creator engine is currently private. Our separate public products demonstrate related first-party software work: Personal Vault, a readable memory/file library, and its desktop client. They are not the source of Creator Flow and not customer traction.
Talk to the team
We welcome a specific story or workflow problem to explore. No payment or subscription is created by contacting us.