LORE · Voice engine
Hosted voice APIs are built for the thirty-second read: a promo, a demo, a line of dialogue. Training is the opposite shape — long, repetitive, revised, and re-rendered every time a standard changes. So LORE narrates with voices we license directly, on hardware we run.
Know the tradeoff
ElevenLabs and the rest of that category solved short-form synthesis extremely well. Nothing below is a knock on the technology — it is a list of the places where a training catalogue and a metered creative tool pull in opposite directions.
| Hosted voice API | LORE | |
|---|---|---|
| Who the voice is | A stock voice from a shared library, or a clone you upload. Your learners may have heard it selling software last week. | Craig and Randi. Professional talent we licensed by name, who know their voice is teaching this material. |
| Consent and provenance | Depends entirely on what the customer uploaded. The platform cannot vouch for a voice you cloned. | Signed talent consent on file for every voice, with the reference recordings and the render provenance kept alongside the course. |
| How cost scales | Metered by the character on a monthly plan. Every revision of every lesson bills again. | A GPU running for an hour. Re-narrating a whole course after a standards update costs machine time, not a bigger tier. |
| Consistency across a course | Tuned per request. Sixteen separate calls can drift in pace and level across a sitting. | One model, one reference, one batch. The bundle is loudness-matched and verified lesson by lesson before anything attaches. |
| Where the narration is made | On the vendor’s servers. Your lesson scripts go with it. | On a GPU we control, against our own model weights. Approved scripts and finished audio stay on infrastructure we run. |
| What you get back | An audio file. Everything after that — hosting, lessons, checks, records — is your problem. | A published course: narration attached to lessons, questions tied to the source, corrections, progress and certificates. |
Why it matters
01
People sit with an instructor for hours. A recognisable, consistent narrator is the difference between a course and a screen reader — and a named, consenting person is the only defensible answer when a safety officer asks whose voice that is.
02
Standards change; courses get re-cut. Under per-character billing every correction is a new invoice, which is exactly the pressure that keeps out-of-date training in service. Our cost is machine time, so re-rendering is a decision about accuracy, not budget.
03
An audio file is the easy part. LORE carries the whole path: extract the source, draft lessons and questions against it, hold everything at a human review gate, narrate, publish, and record who finished.
04
A manual is often the thing an organisation is least willing to paste into a public API. Narration runs on our own hardware against our own weights, and the approved scripts and audio stay on infrastructure we run.
05
Training prose is long, technical and full of citations, units and acronyms. The scripts are written to be heard — and every rendered lesson is checked against its script before it is allowed near a learner.
Side by side
The opening of two lessons LORE built from OSHA 3404, Laboratory Safety Guidance — a public-domain federal document. The script is LORE’s; the voices are Craig’s and Randi’s. Neither clip was touched after it came off the renderer.
“A short task still needs the controls identified in its approved procedure.” Lesson script · drafted from the source, approved by a human
“The paperwork has successfully documented the existence of more paperwork. You still need to know how to do the job.” Lesson script · drafted from the source, approved by a human
Source: OSHA 3404-11R (2011), public domain · licensed voice talent, signed consent on file · rendered on our own GPU
How it runs
Nothing is narrated until a person has read it, and nothing is published until the audio has been checked.
LORE drafts lessons and questions from the source document and holds them at a review gate. Your instructors read, edit and approve. No audio exists yet, because narrating a draft is how a mistake gets expensive.
Approved lessons queue as render jobs. A worker on our GPU claims them one at a time and synthesises each lesson against the reference recording for that voice — same model, same settings, whole course in one sitting, so lesson sixteen sounds like lesson one.
The bundle is checked lesson by lesson — length against the script, loudness matched across the set, provenance recorded — and only then attached to the published course. A render that fails the check does not quietly ship.
When a standard changes, the affected lessons are re-drafted, re-approved and re-rendered. That is a normal Tuesday, not a budget conversation.
Send the manual you’d start with. A pilot is one document, a small group of learners, and a few weeks.
