Feature Roadmap
Everything Meadow could become.
V1 ships a focused, transitional AAC tool for children 12–48 months. This page collects features that didn’t make V1 — some because the scope is intentionally tight, some because the technology isn’t ready, some because they’re genuinely new ideas worth exploring.
Nothing here is committed. Everything here is considered.
The underserved moment
Tier 2 (Word Combinations, 18–24 months) is where a child stops being a button-presser and starts being a communicator. They go from labeling the world — “milk” — to controlling it — “want milk,” “no bath,” “love daddy.” Every AAC app covers this age range, but they all treat it the same way: smaller grid, fewer folders, same paradigm as their adult version scaled down. None of them are designed around what makes this moment magical.
These enhancements make Tier 2 the reason families choose Meadow.
Scene-Tap Micro-Moment
Tier 2
At Tier 1, tapping a scene object speaks its name. At Tier 2, the same tap opens a what about it? moment — the object glows, and the edge words shift to become about that object: “want [milk],” “more [milk],” “no [milk],” “my [milk].”
The child taps the THING (natural — toddlers point at things) and then taps WHAT ABOUT IT. This mirrors how toddlers actually communicate: point at milk, then grunt “more.” Meadow replaces the grunt with a second tap.
This is scene-action synergy at its purest. The scene object becomes half of the sentence. No grid-based app can do this because they don’t have scenes.
“No” as Power
Tier 2
Negation is the most powerful thing a pre-verbal child can learn — it gives them agency. When the child taps “no” + “bath,” the scene reacts. Bath scene objects dim. The app offers alternatives: “No bath. What do you want?”
The child’s words have visible consequences. This teaches cause-and-effect communication — I said something, the world changed. That’s intoxicating for a toddler who has spent their whole life being misunderstood.
Companion Expansion
Tier 2 Tier 3
The companion doesn’t just celebrate. It extends. Child says “want milk.” Companion responds: “You want milk! Yummy milk!” That’s expansion — taking the child’s utterance and adding one element. Every SLP does this in therapy. No AAC app does it automatically.
For the child, the magic isn’t the clinical technique — it’s that the companion heard them. Understood. Talked back at their level. For the parent, the companion is modeling what they should be doing — the Ms. Rachel effect, without a tutorial.
Routine Prediction Chains
Tier 2
Bath time: the scene pre-loads a short flow. “Wash” → companion shows body parts → child taps “wash hands” → “more” or “all done” → “out” or “dry.”
Each step is a 2-word combination the child composes, but the next step is anticipated by the routine. The child isn’t navigating — they’re flowing through a familiar sequence, building real sentences at each beat. Routines are scaffolding for word combinations because the context is so predictable the child can focus on the words, not the situation.
Turn-Taking with the Companion
Tier 2
The companion says “I want…” and pauses, looking at the child expectantly. The child taps “juice.” Companion: “Juice! I want juice too!” Then: “What do YOU want?” Child taps “want” + “cookie.”
This is conversational turn-taking at the 2-word level. The child isn’t just producing language — they’re conversing. With a friend who’s at their level. This is what Speak With Me becomes at Tier 2: not single-word modeling, but combination modeling through conversation.
Feelings Combinations
Tier 2
At Tier 1, tapping “sad” speaks “sad.” At Tier 2, the child can compose: “feel sad,” “mama sad,” “no sad.” Emotional vocabulary + combination = the beginning of emotional intelligence.
The companion responds with empathy: “You feel sad. That’s OK to feel sad.” No AAC app does emotional scaffolding at the word-combination level. Most don’t even have feelings vocabulary that works without navigating three folders deep.
Known gaps, intentionally deferred
V1 ships with intentional constraints — closed vocabulary, English-only, no data export, no switch access. These are scope decisions, not oversights. Each gap has a plan.
Bilingual & Multilingual Support
V2 Critical gap
22% of US children grow up bilingual. Every major AAC competitor offers multilingual support. Meadow V1 is English-only.
V2 adds language packs: vocabulary, voice clips, and scene labels in Spanish (first), then additional languages. The scene-based architecture makes this cleaner than grid apps — scenes are visual, words are the localized layer.
Switch Scanning & Alternative Access
V2 Critical gap
Switch scanning lets motor-impaired children use the app with a single physical button. The app highlights elements one at a time; the child presses the switch to select. Required for many E2510 funded devices.
V1 relies on direct touch. V2 adds linear and group scanning modes with configurable timing, compatible with Apple’s Switch Control framework.
Data Tracking & SLP Reporting
V2 Significant
SLPs need progress data for insurance documentation and therapy planning. What words is the child using? How often? What combinations? Is vocabulary diversity growing?
V2 adds on-device usage tracking with exportable reports. No cloud, no accounts — data stays on the iPad. PDF and CSV export behind the parent gate for SLP sessions.
Custom Vocabulary
V2 Significant
V1 is a fully authored, closed vocabulary. Every word has the full content stack — voice clip, symbol, sentence engine integration, companion engagement prompts. Custom words would be second-class citizens.
V2 adds SLP-configurable custom words with photo symbols and Apple TTS. Custom words inject into the appropriate scene context (e.g., a pet’s name appears in the living room scene).
Gestalt Language Processing
V2 Significant
Gestalt language processors (many autistic children) acquire language in chunks rather than single words — “to infinity and beyond” before “go.” The hottest topic in SLP right now.
Meadow V1 serves analytic language learners (single words → combinations → sentences). GLP support requires architecturally different elements: phrase-level tiles, video-based visual scene displays, echolalic script banks. A meaningful addition, not a bolt-on.
SLP Vocabulary Pack Marketplace
V2
SLPs currently sell custom AAC materials on Teachers Pay Teachers as a workaround. No AAC app has a native marketplace.
Meadow’s world map architecture makes vocabulary packs genuinely rich — a pack isn’t a grid layout, it’s an illustrated scene location with vocabulary, context, sentence engine integration, and therapy targets. SLPs author once; families install into their world map.
Clock & Backpack
V2
Clock: A time-of-day layer on the world map. “It’s breakfast time — go to kitchen?” Supplements spatial navigation with routine-based suggestions. Builds on the time-aware routine glow already in V1.
Backpack: A portable vocabulary the child carries between scenes — favorite words, personal items, things that don’t belong to any one room. Always accessible regardless of which scene is active.
Innovation pipeline
These are ideas, not commitments. Some may never ship. All of them are technically feasible on iPad hardware using Apple’s on-device frameworks. They represent the innovation trajectory that makes Meadow more than an AAC app — it’s a platform for how children learn to communicate.
AR Point-and-Speak
Concept
Child holds the iPad, sees the real world through the camera, taps on recognized objects, and the app speaks the word aloud. Turns the physical world into a vocabulary board.
Fully offline via Apple Vision + CoreML + ARKit. Like Pokémon Go but for language. No competitor has attempted this.
Gamified Sentence Builder
Concept Tier 3+
Drag-and-drop sentence construction as play for older children (4–5+). “Build a salad” style game — drag word ingredients together to construct a sentence. Teaching grammar through play.
Targets children who have progressed past selecting from predictions and are ready for independent sentence construction. The transitional step toward getting off the device.
ShapeSpeak
Concept Tier 3 only
Speech production mini-game. A shape drifts toward a matching cutout; the child says the target word; on-device speech recognition grades the attempt; shape passes through on a close-enough match.
Opt-in only, for children on a verbal trajectory. The first feature that coaches the child’s own voice instead of substituting for it. Not for children who will always use AAC — and that’s fine.
Reverse Storytime
Concept
During a parent’s bedtime read-aloud, the app uses on-device speech recognition to surface matching Meadow symbols in time with the parent’s voice. The child follows along visually — like a reader following text, but in pictures.
Works with any book, no preparation. Piggybacks on a ritual that already exists rather than asking families to add a new one. Words the child hears at night become words they recognize in the app the next day.
What Meadow has that no one else does
V1 ships with six genuine differentiators. The features on this page would extend that lead while closing the gaps competitors will point to.
2. Grammar-aware symbol prediction (market first)
3. Triple modality — picture + speech + ASL on every tap
4. Companion as integrated modeling agent
5. Warm celebration + expectant pausing
6. Transitional design philosophy
2. Switch scanning (motor-impaired access) → V2
3. Data tracking & SLP reports → V2
4. Vocabulary size (700 vs 16,000+) → intentional
5. Custom vocabulary → V2
6. Gestalt language processing → V2
Closest competitor
ChirpBot (launched 2025–2026): free core, AI sentence phrasing, 12 languages, 12mo–12yr target. Their “Word Burst” feature (verb form toggle) mirrors Meadow’s morphological marking. They lack scene-based context, companion modeling, and the transitional philosophy, but they’re the closest direct competitor and should be monitored.