Meadow
Deliverable Review
Incorrect. Try again.
M2 Build Spec  /  M2-013

My Buddy

An always-present guide who teaches words, shows signs, and celebrates — but never talks over the child.
M2-013
v1.0 · May 2026
Milestone: M2 Engine Foundation
Design sources: M1-003 My Guide Interaction Design Voice & Sound Design
01 Meet the Buddy

A warm, animated guide in the corner of every screen

The Buddy is a friendly character that lives in the bottom-right corner of every scene. It gently breathes and bobs, a calm presence that never demands attention but is always ready to help. It is not the child’s voice — it’s the child’s helper.

Think of it as a patient friend sitting beside the child. Not a teacher, not a game character — a warm companion who proactively reaches out when the child might want to learn a new word, shows how to sign it, and celebrates every attempt at communication.

The Buddy is separate from the Avatar, which sits in the bottom-left corner and is the child’s emotional self-expression hub. The Avatar is “me.” The Buddy is “my friend who helps me learn.”

🐻
Always visible
Gently breathing · One tap to engage
πŸ₯›
milk
πŸ₯£
cereal
πŸ§ƒ
juice
🍎
apple
🍌
banana
🚰
water
πŸͺ
cookie
πŸ₯ͺ
sandwich
🀟
cereal
ASL sign
πŸ‘§
My Avatar
🐻
My Buddy
Kitchen scene — Speak With Me active, “cereal” spotlighted, signing bubble from Buddy
02 Speak With Me

A proactive engagement mode that reaches out first

Foundation-tier children aren’t yet initiating communication. Instead of waiting for the child to tap, the Buddy reaches out — spotlighting a word in the scene, saying it aloud with a warm parentese voice, showing the sign, and celebrating any response.

πŸ’‘ 1

Spotlight a word

The Buddy picks a word from the current scene’s vocabulary (weighted toward least-recently-engaged words). Everything else dims gently. The chosen word glows in place — it stays in its scene context, not pulled into a separate UI.

πŸ—£οΈ 2

Present with three modalities

The Buddy speaks the word in its own warm voice (parentese TTS via ElevenLabs), while simultaneously showing the ASL sign in a bubble that appears above it. The child sees the picture, hears the word, and sees the sign — three ways to learn at once.

⏳ 3

Expectant pause

The Buddy waits. This is the most important step — the silence that invites the child to try. Duration adapts by developmental tier: 30–45 seconds for Foundation (youngest children need more time), 10–15 seconds for Explorer, 5–10 seconds for Communicator.

πŸŽ‰ 4

Celebrate any response

There are no wrong answers. The Buddy responds differently based on what happens — but always positively.

🌟

Taps the word

Big celebration — full confetti, Buddy bounces with excitement, the word speaks in the child’s own voice.

πŸ‘

Taps anything

Smaller celebration — gentle sparkle, encouraging Buddy reaction. The tapped word still speaks normally.

πŸ€—

Timer expires

Gentle advance — no negative feedback. Buddy gives a warm nod and moves to the next word. Every session is progress.

➑️ 5

Advance to next word

The Buddy cycles through 3–5 words per session, then returns the app to normal reactive mode. No auto-repeat — the child or parent taps the Buddy again when they’re ready for another round.

Clinical foundation: Ms. Rachel adapted for interactive AAC

Speak With Me adapts six SLP-validated techniques from Ms. Rachel into interactive form: parentese prosody (slower, musical speech), expectant pausing (silence that invites response), triple-modality presentation (picture + word + sign), routine-based scenes, fill-in scaffolding, and warm differential celebration. The difference: Ms. Rachel is passive video. Meadow is interactive — the child’s response matters and is celebrated.

03 Signing Demonstrations

One tap = picture + spoken word + ASL sign

Every word presentation includes an ASL sign demonstration. The sign appears in a bubble that overlaps the Buddy’s position — visually, the sign comes “from” the Buddy. Each sign displays for 1–2 seconds, long enough to register but short enough to keep the rhythm of engagement.

Signs appear during Speak With Me engagement and can also be triggered independently by long-pressing any word tile on screen. This gives parents and SLPs a way to model signing outside of the engagement loop.

🀟
milk
Squeeze motion
πŸ‘‹
more
Fingertips together
βœ‹
all done
Palms turn outward

Why signs alongside speech

Research on total communication shows that pairing speech output with sign language does not delay verbal development — it accelerates it. Children who use signs as a bridge to speech consistently develop verbal language earlier than those who don’t. The Buddy models signs so the child (and parent) can learn them together.

04 When the Buddy Speaks vs. Stays Silent

The child’s voice always takes priority

The most important rule: the Buddy never talks over the child. When the child is communicating, the Buddy is silent. When the child is learning, the Buddy is active.

πŸ”Š

Buddy is Active

The Buddy speaks, animates, and engages during guided learning moments.

  • πŸ’‘ Speak With Me engagement loop
  • 🀟 Signing demonstrations
  • πŸŽ‰ Celebrations and rewards
  • πŸ‘‹ Session greetings and transitions
🀫

Buddy is Silent

The Buddy breathes quietly and does not produce sound or animation that could distract.

  • πŸ‘† Direct word taps (child communicating)
  • πŸ’¬ SCS conversation flow (child is in dialogue)
  • πŸ’› Feelings selection (child expressing emotion)
  • πŸ—ΊοΈ Scene navigation (child exploring)

Why silence matters as much as speech

AAC devices that produce unsolicited output train children to be passive listeners rather than active communicators. The Buddy’s silence during communication acts is a deliberate clinical choice: it reinforces that the child’s voice is the important one. The Buddy helps before and celebrates after, but during the act of communication, it steps back.

05 The Buddy’s Voice

A distinct voice that is clearly “not me”

The Buddy has its own voice, separate from the child’s communication voice. This distinction is critical — the child needs to understand that their speech output (“my voice”) is different from the Buddy talking to them (“my friend’s voice”).

🐻
Buddy’s Voice
πŸŽ™οΈ ElevenLabs Voice Design
🎭 Warm, playful, teddy-bear quality
🎬 Pre-recorded scripted clips
πŸ“Š Three tones: neutral, expressive, urgent
πŸ‘§
Child’s Voice
🍎 Apple TTS Premium
🎯 Parent selects during setup
πŸ”Š Real-time synthesis (any word)
🏠 Same voice for all words

Why two voices

Children with ASD and other developmental differences rely heavily on predictability. The child’s communication voice must be consistent across every word — bundled vocabulary, custom names, sentence constructions. The Buddy’s voice is deliberately different so the child learns: “that warm voice is my friend talking to me; this other voice is me talking.” The contrast is a feature, not a limitation.

06 Acceptance Criteria

What “done” looks like for My Buddy

Every item must pass before the Buddy ships.

Requirement What This Means Status
Visible on every screen Buddy appears in bottom-right corner of every scene, breathing animation active, tappable at all times Required
Speak With Me loop Five-step engagement cycle completes: spotlight → present → pause → celebrate → advance. 3–5 words per session. Required
Three-tier response Taps target word = big celebration. Taps anything = small celebration. Timer expires = gentle advance. No negative feedback ever. Required
ASL signing bubble Sign displays from Buddy position for 1–2 seconds per word during engagement. Long-press trigger outside engagement. Required
Silent during communication Buddy produces no sound or prominent animation during direct taps, SCS flow, feelings selection, or scene navigation Required
Distinct voice Buddy voice (ElevenLabs) is clearly different from child’s communication voice (Apple TTS). Three tonal variants available. Required
Tier-adaptive pauses Expectant pause duration matches developmental tier: 30–45s Foundation, 10–15s Explorer, 5–10s Communicator Required
No auto-repeat Engagement ends after 3–5 words and returns to reactive mode. Child or parent re-initiates by tapping Buddy. Required