Two English-Language AI Hosts.
One Repeatable Content System.
A Smile Zemi-specific concept for persistent English-language personas that can explain the product, guide parents, host videos, and extend into podcast and social content — without requiring a traditional on-camera spokesperson.
Concept personas shown on this page are AI-generated and are not real Smile Zemi employees or customer testimonials.
Don't Build 15 Videos. Build Two Hosts.
This persistent host system is a more valuable asset than the initial videos or podcast. Once the characters, voices, editorial lanes, and visual identities are established, Smile Zemi can reuse them across parent education, feature launches, onboarding, YouTube, social, FAQs, homeschool content, and future podcast episodes.
Two Editorial Lanes, One System
Each host is built to full Signal Lab persona architecture — background, voice engine, visual specification, delivery pattern, and guardrails — so scripts stay consistent across dozens of pieces of content.
Emily Carter
The Parent GuideEmily translates Smile Zemi features into the question parents actually care about: "What does this mean for my child?" Warm, practical, reassuring, and clear. Best for parent onboarding, FAQs, feature explainers, product launches, and family-facing content.
Marcus Bennett
The Learning CoachMarcus explains why a learning approach matters. Clear, curious, grounded, and more analytical than Emily. Best for curriculum explainers, feature demonstrations, homeschool content, longer-form YouTube, and podcast episodes.
Why Handwriting Still Matters
Hosted by Emily Carter · 60–75 second finished video. A parent-first explanation of how a digital learning experience can still ask children to write, listen, solve, and practice — turning "screen time" into active learning time.
Finished sample video — hosted by Emily Carter.
Read the approved 60–75 second script
Parents hear a lot about screen time.
So here's something interesting about Smile Zemi.
Your child isn't just tapping answers on a tablet. They're writing.
When they're learning vocabulary, they can practice spelling words by hand. When they're solving math problems, they're encouraged to actually work through the problem instead of simply choosing an answer.
Why does that matter?
Because learning isn't just about seeing the right answer. It's about doing the thinking that gets you there.
Smile Zemi combines the convenience of a digital learning system with activities that still ask kids to write, practice, solve and respond. And because the tablet can provide feedback as they work, children can keep moving without needing a parent sitting beside them for every question.
That's really the idea behind Smile Zemi: give kids the tools to practice independently — while giving parents confidence that the time they're spending on a screen is actually learning time.
The Persona Is More Than a Face.
Each host is developed with a persistent voice specification, emotional baseline, cadence, delivery pattern, approved phrases, visual rules, and editorial lane. That means the same character can remain recognizable across dozens of pieces of content.
Her full voice is featured in the finished video above.
↑ Watch the sample- Medium-low register, articulate, calm and grounded — conversational, not professorial
- Pace 140–150 WPM, with a deliberate pause before contrast statements
- Emotional arc: curious engagement → analytical clarity → confident takeaway
Finished intro sample — hosted by Marcus Bennett.
How a Persona Is Built
Emily and Marcus aren't a headshot and a script prompt. Each is built to the same structured architecture as every persona in the Signal Lab library — the depth is what keeps a character consistent across dozens of pieces of content instead of drifting piece to piece.
See the full architecture — Emily Carter example
Background — age, hometown, family context, education, and career path, built to give her an internally consistent point of view rather than a generic on-camera presence.
Philosophy — "The best learning tools help children do more on their own without leaving parents in the dark."
Values — Independence, Clarity, Usefulness, Confidence, Balance — the filter every script is checked against.
Narrative style — a repeatable five-step delivery pattern: name the parent concern → introduce the feature → show what the child does → explain why it matters → close with a practical takeaway.
Voice engine — a written ElevenLabs voice prompt plus tuned settings (Stability 0.62 · Similarity Boost 0.74 · Style 0.38), so the same voice identity reproduces across every script.
Emotional arc — a Hume baseline of "warm confidence," mapped beat by beat — curious and relatable at the open, clear through the explanation, reassuring and confident at the close.
Visual / avatar specification — a full portrait and on-camera prompt covering framing, lighting, gesture, and energy level, so the avatar stays visually consistent video to video.
Scripts — approved 30-, 60-, and 90-second models built from the same narrative arc, so new scripts can be generated quickly without losing her voice.
Platform fit — parent onboarding, FAQs, feature explainers, product launches, homeschool content, and podcast — mapped in advance so new requests have a clear home.
Built on an Operating System — Not a One-Off AI Experiment
Emily and Marcus run on the same pipeline already producing this library, live today.
Recurring Shows, Not One-Off Clips
Two branded, AI-hosted podcasts already running in production on a weekly schedule — live shows, not concepts.
Miles From Nowhere
A branded storytelling podcast built around the people and communities behind America's infrastructure buildout. Host Mason Reed is disclosed as an AI-generated persona.
Listen on Apple PodcastsHot Wired
A recurring show built around the skilled-trades audience, using a consistent format and voice to educate and engage listeners. Host Connor Walsh is disclosed as an AI-generated persona.
Listen on Apple PodcastsPersona → Script → Voice → Video → Podcast → Asset Package
The same pipeline that produces Emily and Marcus runs every persona in the Signal Lab library today.
A System That Outlasts the Launch
Once Emily and Marcus exist as fully specified hosts, they don't expire when the first 15 videos ship. The same voice, visual identity, and editorial lanes extend into new feature launches, seasonal campaigns, homeschool content, and future podcast episodes — without rebuilding a persona from scratch each time.
See More Before September 17
This page focuses on Emily, Marcus, and the Smile Zemi concept specifically. For the full scope of what's running in production today — across campaigns, associations, and other partners — the links below go deeper.