AI Presenters

A presenter on screen, no studio required

Some demos need a face. Not because the product is unclear, but because a person on screen holds attention for longer and makes an unfamiliar product feel like it came from somewhere real. Booking that shoot is the expensive part. This is not.

PRESENTERS · 180 LICENSED
In transit
1,284
+12%
On-time rate
96.4%
+2.1%
Avg. dwell
3.2h
−18m
At risk
23
+4
Shipment
Route
ETA
Status
SHP-4821
Rotterdam → Chicago
2 days
On time
SHP-4822
Shenzhen → Long Beach
9 days
Delayed
SHP-4823
Hamburg → Newark
4 days
On time
SHP-4824
Busan → Oakland
12 days
At risk
Maya · Explainer
+176
180
Licensed presenters
34
Languages, lip-synced
6 min
Footage to train your own
+41%
Median completion vs. no presenter

Cast rather than shoot

180 presenters across ages, regions and registers, each licensed for commercial use with no per-video fee.

Four ways to appear

Circle bubble, rounded card, full-frame intro, or a presenter who hands off to the screen and returns for the close.

Consent-gated custom avatars

Train a presenter on yourself or a colleague with recorded consent, an audit record, and revocation on demand.

Stock presenters

Cast a presenter the way you would cast an actor

Filter by language, apparent age, register and setting, then preview each candidate delivering your own opening line. Every presenter is licensed for unlimited commercial use — no per-video royalty, no “ask before you run paid media”, no clause that stops you using them in a market you expand into next year.

  • Filter by language, region, register, wardrobe and setting
  • Preview on your actual script before committing
  • Unlimited commercial use worldwide, included in your plan
  • Consistent presenter across a whole video series
PRESENTERS · 180 LICENSED
In transit
1,284
+12%
On-time rate
96.4%
+2.1%
Avg. dwell
3.2h
−18m
At risk
23
+4
Shipment
Route
ETA
Status
SHP-4821
Rotterdam → Chicago
2 days
On time
SHP-4822
Shenzhen → Long Beach
9 days
Delayed
SHP-4823
Hamburg → Newark
4 days
On time
SHP-4824
Busan → Oakland
12 days
At risk
Maya · Explainer
+176
Custom avatars

Train a presenter on your founder, once

Six minutes of guided footage — a few sentences, a few expressions, a couple of gestures — produces a presenter that looks and moves like the real person, and speaks every language in their own voice clone. Useful when the founder is the story and the founder does not want to re-record the pitch for each of eleven markets.

  • Six minutes of guided capture, shot on a laptop
  • Pairs with a voice clone for a complete likeness
  • Speaks all 34 languages with matched lip movement
  • Recorded consent, audit trail, and revocation by the subject
Composition

Present without covering the thing you are presenting

The presenter is aware of the frame. When a zoom moves to the top-right, the bubble slides to the bottom-left. When a modal opens under the presenter, it shrinks. When the narration is describing something specific on screen, it can fade out entirely and return for the conclusion. You can override all of it with a keyframe.

  • Collision-aware placement against zoom regions and UI
  • Automatic fade-out during detailed on-screen explanation
  • Full-frame intro and outro with a screen-share handoff
  • Per-scene manual placement when you want control
Specifications

The technical detail

Stock library
180 presenters, commercially licensed worldwide
Custom training
6 minutes guided capture; model ready in ~2 hours
Languages
34, with lip movement matched to the target audio
Resolution
Up to 4K, alpha channel available for compositing
Consent
Recorded consent required; revocable with audit trail
Security signed off in a week. Private rendering in our own VPC, no training on our frames, SOC 2 report on request. That is not a sentence I get to write about most AI vendors.
7 daysTo security approval
TATom AldridgeStaff Security Engineer · Halcyon
Questions

About ai presenters

Be honest with yourself about the use case. For a thirty-second intro and a fifteen-second close around a screen recording, presenters read as natural to nearly everyone. For eight unbroken minutes of a talking head, some viewers will notice. We design for the first case and tell you plainly that the second is harder.

Often not. A tight voiceover over a well-directed screen recording converts extremely well and is faster to produce. Presenters help most on top-of-funnel content, founder-led narrative, and localized markets where a local-language presenter builds trust you cannot get from a voice alone.

Yes. Revocation is a right of the person, not the workspace. Revoking disables the avatar for all future renders immediately and flags existing videos that use it so you can re-render them with a stock presenter.