Name: AlloFlow
Author: Aaron Pomeranz

You do not need to understand all of AlloFlow first.

Choose the job closest to yours. Each route explains source requirements, AI needs, and the best place to launch.

Educator: find a classroom tool

Search STEM labs, literacy supports, planning tools, assessment workflows, and prepared resources by the task at hand.

Open the tool finder → 2

Accessibility team: remediate a document

Understand the source-bound pipeline, evidence package, portable Skill, local connector, and required human review.

Explore remediation → 3

District evaluator: choose an access path

Compare Gemini Canvas, the hosted browser app, Desktop, portable Skills, and local connector responsibilities.

Compare paths → 4

Tester or contributor: shape what improves

Try one focused workflow and share an idea, bug, accessibility barrier, or de-identified classroom observation.

Share feedback safely →

Start with the tool you need

Search by classroom goal, subject, or feature. Results explain whether a tool needs source material or AI before you enter AlloFlow.

Source-aware:Some tools start immediately. Others need source text or analyzed source material; every catalog card tells you which.

Browse the full tool finder →

Create accessible document alternatives with traceable evidence.

AlloFlow can analyze a source, rebuild semantic structure, verify exact outputs with multiple engines, and package artifacts for human review. It does not promise automatic legal conformance.

Inspect the pipeline

Follow extraction, semantic rebuild, bounded repair, independent HTML checks, tagged-PDF output, and review.

Use a portable Skill

Download a versioned package whose scripts do not call an AlloFlow server, with clear host and local-tool boundaries.

Plan the right edition

Understand what is local, what needs Gemini or another provider, and how the Claude Desktop connector installs from a checksummed release.

Title II planning dates are approaching. The standards exist to make digital services usable by people with disabilities—not merely to satisfy a deadline.

Read the standards context

Try one focused workflow and tell us what happened.

Short, de-identified feedback about a tool, bug, or accessibility barrier is more useful than a generic review.

Choose a safe feedback route

Documented Features

Registered STEM IDs

SEL Tools

Language Resources

Tutorials

Walkthroughs of the real workflow, start to finish.

                                 gemini.google.com · localhost:3000
                            

Resource Generation

Build a differentiated lesson from a single source · 12 min

                             For Administrators
                        

Administrator Overview

See how AlloFlow supports district-wide goals

Learn More

Short explainer videos for every stakeholder — teachers, families, and multilingual learners.

Watch Overview

For Teachers

See how AlloFlow generates differentiated lessons, aligned assessments, and interactive activities — replacing hours of prep with a single free AI tool.

Watch Overview

For Parents & Families

How families can use AlloFlow at home for reading practice, phonics games, adventure mode stories, and homework support — for less than $8/month.

Watch Overview

Multilingual Learners

Supporting multilingual learners with mirrored language packs, bilingual scaffolds, and AI-assisted translation workflows — not just Google Translate overlays.

🎧 Audio for Families

Kitchen Table Scripts & Gamified Learning

A quick listen for parents and families — discover how AlloFlow turns your kitchen table into a fun, AI-powered learning station with gamified phonics, math adventures, and interactive stories.

Duration: ~5 min • Best enjoyed with a cup of coffee ☕

Try One Tool Right Now

No signup, no install, no account. Three starting points for the three questions educators arrive with: can it fix what I already have, can it help me run my work, and will students actually engage with it.

Fix a document you already have

Take an inherited PDF or worksheet through the remediation pipeline: audit, rebuild, human review, and an evidence report you can keep.

Educator Growth & Evaluation

Walkthroughs, the full observation cycle, and growth trends in a private workspace that stays on your device. For leaders and coaches.

Water Cycle

Ride one droplet through evaporation, condensation, and precipitation, making choices along the way. One of 140+ interactive labs.

Browse the full tool catalog, read the teacher guide, or start with the chapter for school leaders. Evaluation has its own user manual.

Three Pathways. One Mission.

Run instantly in Gemini Canvas, deploy a district-controlled Firebase site, or use AlloFlow Desktop for local files, local providers, and same-room Desktop LAN sessions. The Docker School Box stack is an optional server/appliance path for schools that need district-owned infrastructure.

Choose Your Path

Cloud: Paste the Canvas link. Firebase: Deploy static hosting. Desktop: run the local-first app. Server: use Docker only for optional School Box infrastructure.

Configure & Create

Complete the Quick Start wizard or ask AlloBot to auto-generate a full lesson pack.

Differentiate

Generate quizzes, games, leveled text, and more. Export or go live instantly.

Eliminating the "Differentiation Tax"

Educators face a cognitive burden: adapting one curriculum for dozens of unique learning profiles.

For Teachers: Speed

Manual differentiation takes hours. AlloFlow automates text leveling, glossary creation, and assessment design in seconds.

Teacher Paced Mode for guided instruction
Automated QTI/Canvas Exports
Integrated Answer Keys & Rubrics

For Students: Access

Students receive a personalized "Lesson Pack" and review material at their own pace with built-in accommodations.

Offline HTML Bundles for homework equity
Native Text-to-Speech & Phonics
Gamified Review (XP & Leveling)

Explore the Pedagogy

Dive deep into the research behind AlloFlow. Listen to our AI podcast overview or tap through the interactive slide decks.

The "Why" Behind AlloFlow

Audio Overview • 8 mins

Equity First

Local-First Architecture

AlloFlow is built around a local-first app shell. While AI generation usually requires a provider connection, exported project and HTML bundles can preserve many study resources for offline or low-connectivity use after the materials are created.

🌐 Online: AI Generation

Features that generate new content—like Adventure Mode worlds, DOK Leveled Quizzes, and Text Rewriting—require an active connection to the Gemini Canvas API.

⚡ Offline: Interactive Play

After assets are loaded, many STEM, game, Word Sounds, and Immersive Reader interactions can run locally in the browser. AI generation, cloud providers, and hosted-session features remain connection-dependent.

📥 The Bridge: HTML Bundles

Teachers can generate lessons online, then click "Export to HTML". This bundles the entire interactive experience into a single file that students can play completely offline.

0* Server round-trips for offline core play

Core Core interactions offline

Differentiation

Instant Leveled Text

Adapting a single reading passage for five different reading levels previously took hours. AlloFlow's Bimodal Generative Pipeline reconstructs raw texts dynamically — one click, multiple outputs.

🔬 Syntax & Semantic Downgrading

The AI doesn't just swap synonyms — it restructures compound-complex sentences into simple S-V-O patterns while preserving the narrative "Golden Thread" across all levels.

🧵 Golden Thread Preservation

Every leveled version retains the same core narrative, characters, and key vocabulary so all students can participate in the same class discussion.

📚 Dolch Word DB Integration

Lower levels automatically substitute vocabulary from the Dolch high-frequency word list, ensuring emerging readers encounter sight words they're actively learning in class.

5 Reading levels from one source

<60s Generation time per level

Engagement

Multi-Modal Workflows

6+ output formats from one standard

One size never fits all. AlloFlow uses structured content representations so source material can be reviewed and rendered into different UIs—gamified adventures, interactive quizzes, printable worksheets, and more.

Deterministic HTML Generation

AI outputs are rigorously mapped into predictable data structures, allowing an Escape Room puzzle and a 3D Dice Adventure to stem from the exact same mathematical standard.

Zero-Code Authoring

Teachers toggle modes in a visual UI—no HTML, CSS, or JavaScript knowledge required. The engine handles all rendering automatically.

📋 Live Preview Dashboard

Teachers see a real-time preview of each modality before distributing to students—3D Adventure, Escape Room, Quiz, Comic, or Worksheet—all rendered from the same source data.

Science of Reading

Phonemic Precision

44 IPA phonemes mapped to orthography

English spelling frequently violates spoken rules. The Word Sounds Studio bridges orthography with auditory IPA sounds through synchronized playback, aligned with the Science of Reading's phonemic awareness pillar.

Advanced Audio Orchestration

To prevent TTS "ghost loops", audio strings are cached in proxy layers and sequenced perfectly to prevent overlapping reads during rapid student interactions.

Elkonin Box Integration

Interactive sound boxes let students tap individual phonemes while seeing the IPA symbol and hearing the isolated sound—bridging the gap between letters and speech.

🎯 Multi-Sensory Strategy Alignment

Visual letter highlighting, auditory IPA playback, and kinesthetic tap-to-segment actions combine three modalities in a single exercise—aligning with Orton-Gillingham and structured literacy best practices.

3 Modalities per exercise

5 Phonemic awareness activities

Scaffolding

Interactive Tour Engine

5 step guided walkthrough sequence

Allobot is an embodied pedagogical agent. Rather than passively chatting, it parses JSON to physically manipulate the DOM, guiding students visually to exactly what they need—reducing cognitive load for learners who struggle with navigation.

data-help-key Binding

Every element is tagged. The AI generates deterministic coordinates for the Bot Avatar to scroll to, open drawers, and highlight targets in sync with Text-To-Speech.

Multi-Sensory Navigation

Visual highlighting, spoken narration, and animated scrolling combine to guide students through features—fulfilling UDL's multiple means of representation.

♿ Accessibility-First Navigation

Screen-reader-compatible ARIA labels are auto-injected by the tour engine. Every highlighted element receives focus management and keyboard-navigable controls—aligning with RTI Tier 1 scaffolding.

100% Keyboard accessible

ARIA Auto-injected labels

FERPA-Aligned Data Paths

Privacy-Minimizing Architecture

0 student PII required by default

AlloFlow is designed to minimize student data collection and keep ordinary work in the browser, downloaded files, Desktop runtime, or school-controlled infrastructure when those paths are selected. Cloud AI, hosted live-session paths, provider terms, and retention settings remain deployment-dependent; districts should review their contracts and configuration.

Local Context Compression

All quiz scores, XPS, and RTI metrics are heavily compressed (via LZ-String) and persisted explicitly inside the district's local browser context.

FERPA/COPPA Support Depends on Deployment

AlloFlow can support FERPA-aligned school-controlled deployments, but final compliance depends on district configuration, contracts, retention policy, and actual classroom use.

🔐 Minimized Data Exposure

Local-first paths can keep work on teacher devices or school-controlled infrastructure. Cloud and hosted paths may involve provider or server processing, so districts should review deployment boundaries, contracts, and retention settings.

Varies Provider / hosting agreements

LZ String-compressed storage

Community

Empowering Families

$0 AlloFlow software license cost

Response to Intervention (RTI) thrives when families are involved, but high-tier educational software is usually gatekept by district licensing fees. AlloFlow lowers this barrier with free software access for families, 63 mirrored language-pack files, and AI-assisted content translation.

💰 The $0 Software Extension

Parents can practice identical phonics strategies and manipulatives at the kitchen table that teachers use in the classroom, ensuring continuity of care across all settings.

🔄 Consistent Strategies Across Settings

The same science-of-reading decodables, leveled passages, and interactive quizzes students see in school are available at home—no separate app or subscription required.

🌐 Multilingual Family Support

Parents and guardians can view all lesson content in their native language via the built-in translation layer—bridging the school-to-home gap for multilingual households without external tools.

63 mirrored language-pack files

$0 AlloFlow software license cost

STEM & Science

Interactive STEM Lab

65 interactive tools spanning 9 disciplines

A full virtual lab covering math, advanced math, life science, earth & space science, physics & chemistry, computer science, creative design, social studies, and applied simulation. Students manipulate real simulations — not passive videos.

🔬 Virtual Lab Equipment

Circuit builders, wave visualizers, DNA labs with CRISPR simulation, titration labs, flight simulators, fire ecology, epidemic modeling, semiconductor physics, and a coding playground — all running in the browser with zero install.

🐠 Immersive Simulations

Explore a virtual aquarium with 30+ AI-driven species, a human anatomy explorer with 129 structures across 10 body systems, a neuroscience atlas with 30+ brain regions, a pharmacology lab with 20+ neurotransmitter pathways, and a geology collection with 30+ mineral specimens.

📐 Cross-Curricular Coverage

From K-2 base-ten blocks to high school function graphers, chemical equation balancers, and data plotting tools. Aligned to NGSS and Common Core Math standards.

⚡ CDN-Powered Loading

The entire STEM Lab loads as a single external module via jsDelivr CDN. Zero impact on the core app's performance — tools are fetched only when selected by the user.

5 Science Disciplines

0 Install required

Clinical & Behavioral Support

BehaviorLens Clinical Suite

18+ clinical tools across 8 categories

A comprehensive ABA and behavioral assessment toolkit built for BCBAs, school psychologists, and special educators. Every tool runs locally — no student data ever leaves the device.

📊 Data Collection & Analysis

IOA calculators (5 methods including exact agreement, total count, and interval-by-interval), DTT data sheets, interval recording, frequency/duration/latency records, and cumulative graphing.

🎯 Clinical Instruments

Preference assessments (MSWO, Paired, Free Operant), task analysis with chaining and prompt hierarchies, token economies, behavior contracts, DRO/extinction protocols, and treatment integrity tracking.

🤖 AI-Powered Tools

Natural language ABC analysis, IEP goal generators, observation coach, progress reports, caseload dashboard, MTSS tier manager, and effect size calculators (Tau-U, NAP, PND) — across AI-assisted workflows.

18+ Clinical tool types

FERPA Deployment-dependent privacy

Clinical Reports

Report Writer

10 step wizard with AI accuracy verification

A psychoeducational report wizard purpose-built for school psychologists. Triangulated generation with self-healing accuracy verification ensures every claim is traceable to your original data — the safest AI-assisted clinical report tool that exists.

🔒 Fact-Chunk Accuracy Pipeline

Scores are extracted as immutable facts. The AI generates N parallel drafts, selects the most evidence-grounded one, then runs score-text verification, cross-section consistency checking, and a self-healing dual-pass audit that auto-fixes errors and re-verifies.

📝 17 Assessment Presets + Multi-Format Export

WISC-V, WIAT-4, BASC-3, Vineland-3, BRIEF-2, Conners-4, and 11 more. Generate reports at technical, parent, or student reading levels. PII is scrubbed from all AI-facing pipelines with developmental norms cross-referencing (attention span, tantrums, social play, vocabulary by age).

PII Auto-scrubbed from AI pipelines

3 Reading levels per report

Built-In Research

Research Suite

1-Click study setup with automated data collection

AlloFlow includes a complete clinical research infrastructure built directly into the platform for core study setup and collection; external analysis tools remain optional. Design a study, collect data, and export CSV files for analysis.

🔬 Research Mode

Name your study, set an IRB number, and configure auto-survey frequency (every 3/5/10/15 sessions). Research Mode automatically logs configured sessions automatically after setup.

📊 Embedded Surveys & CBM

Built-in student, teacher, and parent/guardian surveys launch from the Research Tools panel. Import external CBM scores (DIBELS, AIMSweb) to link with AlloFlow usage data for correlational analysis.

📈 Probes & Progress Monitoring

Math fluency probes (computation accuracy + speed), Oral Reading Fluency with AI-scored WCPM and accuracy heat maps, and literacy fluency probes — all with exportable, timestamped data.

CSV One-click export

0 External tools needed

Developer Ecosystem

Modular Architecture

0 lines of backend code required for new modules

AlloFlow's "app-within-an-app" framework lets developers build independent educational modules that inherit the full AI infrastructure — translation, Gemini access, data storage — with zero backend code.

🔌 Hub-and-Spoke Model

The parent app provides core infrastructure (Gemini AI, translation, device sync). Community modules plug in as lightweight single-file apps with full creative freedom over UX and logic.

📦 CDN-Based Off-loading

Modules are hosted on GitHub and served via jsDelivr CDN. Only the modules a user selects are loaded — ensuring scalability even with 50,000+ lines of community-contributed code.

🤖 Agentic Development

This architecture is optimized for AI-assisted "vibe coding." Developers can use agents like Antigravity to build robust tools on top of the existing, battle-tested AlloFlow foundation.

AGPL Open-source modules

∞ Extensible modules

Gamification

Adventure Mode

5 narrative rooms per quest with boss battles

Reading becomes an interactive RPG. Students navigate through themed rooms, answer comprehension questions to unlock doors, collect inventory items, and face boss battles — all generated from any lesson content.

🎲 3D Dice & XP Progression

Physical dice rolls determine battle outcomes. Students earn XP, level up character portraits, and unlock achievements — creating intrinsic motivation to read deeply.

📖 Storybook Export

Completed quests export as illustrated storybooks that students can share. Each page recaps their journey, choices, and discoveries — turning reading into a creative artifact.

🏰 Dynamic Room Generation

Gemini AI generates themed rooms, NPC dialogue, obstacle logic, and boss encounters from any standard or reading passage — with reduced setup time after the lesson content is configured.

RPG Reading engine

Reduced Manual prep after content setup

Multilingual

63 Language-Pack Files

63 mirrored language-pack files

Multilingual learners deserve tools in their home language. AlloFlow mirrors UI keys across the current language-pack files and supports AI-assisted content translation. Native review is still important for institution-facing use.

🌐 Full UI Translation

Every button, menu, tooltip, and instruction translates automatically. Students interact with the entire app in their native language, not just the lesson content.

📝 Bilingual Views

Side-by-side English + home language display lets MLL students read content in both languages simultaneously, building English skills while retaining comprehension.

🗣️ Culturally Responsive

Teachers can override language at the activity level. A Spanish-speaking student can read English content with Spanish scaffolds — all within the same lesson.

1 Click to translate

RTL Script support

Engagement

Interactive Game Modes

8+ game types generated from any lesson

Every lesson can become a game. AlloFlow generates Review Games, Bingo, Memory Match, Crossword, Word Search, Escape Room, Team Showdown, and Concept Sort — all from the same academic content, with minimal setup after the content is selected.

🎮 Zero-Code Game Generation

Gemini AI converts any reading passage, vocabulary list, or standard into ready-to-play games. Teachers select a game type and the content is mapped automatically.

🏆 Escape Room Engine

Teams solve 6 puzzle types across themed rooms with clue chains, streak bonuses, lives, hints, and a Final Door synthesis challenge — gamified comprehension at its peak.

📊 Team Showdown

Firebase-powered live competition with real-time leaderboards and team scoring. Teachers control rounds from their device while students compete on theirs.

6 Escape Room puzzles

Live Multiplayer support

Computer Science

Gemini Coding Bridge

3 sandbox environments: Python, React, p5.js

Gemini AI auto-generates sandboxed coding environments directly from lesson content. A reading passage about ecosystems becomes a p5.js simulation; a math lesson becomes a Python calculator.

💻 Instant Sandbox Generation

One click converts any lesson into a live coding environment with pre-written starter code, instructions, and expected outputs — ready for students immediately.

🎨 Cross-Curricular Coding

Art becomes p5.js sketches. Math becomes Python functions. Science becomes data visualizations. CS integration happens naturally within existing curriculum.

🔒 Safe Execution

All code runs in sandboxed iframes with no access to the parent application, student data, or external networks. Students experiment freely without security risk.

AI Generated code

🔒 Sandboxed execution

RTI Monitoring

Oral Fluency Coach

WCPM calculated automatically from speech

Students read aloud and the AI calculates Words Correct Per Minute (WCPM) with a color-coded accuracy heat map. Results feed directly into RTI progress monitoring dashboards.

🎙️ Speech-to-Text Analysis

Students read a passage aloud using their device microphone. The AI transcribes and compares against the original text, identifying omissions, substitutions, and self-corrections.

🗺️ Accuracy Heat Map

Each word displays a color-coded overlay: green for accurate, yellow for hesitation, red for errors. Teachers instantly see fluency patterns without manual running records.

📈 RTI Data Integration

WCPM scores are automatically logged to the local progress monitoring dashboard. Track growth across sessions, set benchmark goals, and export data for IEP documentation.

0 Manual running records

RTI Progress monitoring

Collaboration

Democracy Mode

Live class voting with group-level differentiation

Students vote on story directions, vocabulary choices, and lesson paths in real time. The system then pushes different leveled content to different groups based on their votes and reading levels — true democratic learning.

🗳️ Live Classroom Voting

Students vote from their devices on story choices, discussion prompts, or activity types. Results display live on the teacher's projected screen with animated vote tallies.

🔀 Group-Level Differentiation

After voting, the system automatically pushes different content to different groups. Advanced readers get enrichment while emerging readers get scaffolded versions — same lesson, different paths.

🤝 Student Agency

Students own the direction of their learning. Choices create investment. Investment creates engagement. Engagement creates retention. Democracy Mode closes the loop.

100% Student voice

Auto Group differentiation

Accessibility

Immersive Reader

WCAG 2.2 AA-oriented accessibility evidence

Built-in Immersive Reader with Dyslexia-friendly fonts, line focus masks, syllable splitting, grammar color-coding, and picture dictionary — all students read independently and comfortably.

🔤 Dyslexia-Optimized Fonts

OpenDyslexic and other specialized typefaces increase letter spacing, weight bottoms of characters, and reduce visual confusion — toggle-able per student without affecting others.

📏 Line Focus & Reading Masks

Configurable reading ruler highlights 1, 3, or 5 lines at a time while dimming surrounding text. Reduces visual overwhelm for students with attention or tracking difficulties.

🎨 Grammar Highlighting

Color-codes nouns (blue), verbs (red), adjectives (green), and adverbs (gold) in any text. Syllable splitting breaks long words into readable chunks with visual separators.

♿ Keyboard + SR parity

4 Parts of speech

Assessment

Exit Tickets & Quiz Engine

4 DOK levels with auto-tuned question complexity

Gemini AI generates standards-aligned multiple-choice questions at the teacher's chosen Depth of Knowledge level. Export directly to Canvas in QTI format or use as in-app exit tickets with instant grading.

📝 AI-Generated Questions

From any passage, Gemini generates DOK 1 (recall), DOK 2 (skill), DOK 3 (strategic thinking), or DOK 4 (extended thinking) questions — teachers choose the cognitive demand.

📤 Canvas QTI Export

Export as QTI 2.1 for Canvas, Schoology, or other compatible LMS workflows; verify import settings. Question banks build automatically over time.

✅ Instant Grading & Feedback

In-app exit tickets grade instantly with wrong-answer explanations. Teachers see class-wide results in real time — no paper, no scanning, no data entry.

QTI Canvas-ready export

0s Grading time

Analytics

Teacher Dashboard

3 RTI tiers with automatic progress monitoring

The Teacher Dashboard provides real-time class analytics, individual learner profiles, and RTI/MTSS progress monitoring — all stored locally when using local-first paths. Track growth, set goals, and export IEP documentation.

📊 Class-Wide Analytics

See class averages, score distributions, and completion rates across all activities at a glance. Identify struggling students and celebrate growth — powered by local data only.

👤 Individual Learner Profiles

Drill into each student's journey: quiz scores, fluency rates, phonics mastery, and XP progression. Data-driven insights with reduced cloud persistence when configured for local use.

📈 RTI/MTSS Progress Monitoring

Track students across Tier 1, 2, and 3 interventions with benchmark goals and trend lines. Generate graphs for IEP meetings directly from local data.

Local Data stored on device

IEP Export-ready reports

Mission

The Open-Source Commitment

100% open-source software

Built by a public school educator to dismantle the "Differentiation Tax"—the unacceptable reality that teachers spend their own money on accommodations. AlloFlow is licensed under AGPL v3 and currently has no AlloFlow software license fee. Hosting, hardware, support, and provider usage can still carry costs.

Fully Open Infrastructure

AlloFlow is completely open-source. School districts or independent developers can fork, modify, and host their own internal variants of the platform without vendor lock-in.

No AlloFlow License Fee

Core software is available under AGPL v3 with no AlloFlow license fee; districts still budget for deployment, support, hosting, hardware, and provider usage when applicable.

🤝 Community Contributions Welcome

Educators can submit feature requests, custom games, and lesson packs via GitHub. Every contribution is reviewed and merged to benefit the entire community—turning teachers into co-developers.

AGPL AGPL v3 licensed

$0 AlloFlow software license fee

AI Generation

AI Research Engine

3-Pass research, outline, and citation pipeline

AlloFlow's Source Generator uses a multi-pass AI pipeline with optional Google Search grounding. Content is researched, outlined, written section-by-section, and fact-checked — producing citation-backed lessons from any topic.

🔍 Google Search Grounding

Toggle "Verify Facts" and Gemini searches the live web before writing. Sources are cited inline with numbered references — ensuring lesson accuracy for science, history, and current events content.

📄 Multi-Pass Writing Pipeline

Pass 1 researches and gathers sources. Pass 2 builds a structural outline. Pass 3 writes each section with smooth transitions and embedded citations — producing publication-quality educational content.

🎯 Teacher-Controlled Parameters

Teachers set topic, tone (Informative, Narrative, Persuasive, Humorous), grade level, standards alignment, vocabulary targets, and word count — the AI handles the rest with no prompt engineering required for presets.

5 Tone presets

K–Grad Level range

Literacy

Spelling Bee & Word Games

4 phonemic game modes with voice input

The Word Sounds Studio transforms phonics practice into interactive game modes — Spelling Bee with microphone input, Sight & Spell orthography drills, Elkonin sound boxes, and Word Sorting — all aligned to the Science of Reading.

🎤 Voice-Powered Spelling Bee

Students hear a word via TTS, then spell it aloud using their microphone. Speech-to-text validates each letter in real time, with streak tracking, difficulty progression, and instant corrective feedback.

👁️ Sight & Spell Orthography

Students see a word briefly, it disappears, and they must reconstruct it from scrambled letter tiles. AI-generated distractors ensure phonetically plausible wrong answers — testing true orthographic memory.

💡 Adaptive Hint System

Progressive hints reveal one letter at a time without giving away the full answer. The system tracks which phoneme patterns students struggle with and adjusts future word selection accordingly.

🎙️ Voice recognition

SoR Aligned activities

Visual Learning

Organizers & Decodable Readers

6+ organizer types with AI-generated decodable readers

Transform any lesson into visual thinking tools — Venn diagrams, KWL charts, cause/effect maps, timelines, concept webs, and T-Charts. Plus, generate Science of Reading-aligned decodable readers with controlled phonics patterns.

🗺️ AI-Generated Visual Organizers

Gemini analyzes lesson content and automatically populates organizer fields. Students can edit, expand, and rearrange — building comprehension through visual structure rather than passive reading.

📖 Decodable Reader Generator

AI creates phonics-controlled stories targeting specific letter patterns (CVC, CVCe, blends, digraphs). Each story uses only previously taught patterns — true Science of Reading alignment for emerging readers.

🎓 UDL Multiple Representations

The same content appears as text, visual organizer, decodable story, and interactive quiz — giving every learner a pathway that matches their cognitive strengths and developmental stage.

6+ Organizer types

SoR Decodable readers

Self-Hosted AI

Your AI, Your Hardware

$0/yr Cloud API fees in local mode

AlloFlow supports model-agnostic AI backends. AlloFlow Desktop is the normal local-first teacher-laptop path. The Docker School Box stack is optional server/appliance infrastructure for schools that need district-owned local services.

🐳 Optional Docker School Box

docker compose up -d launches local services such as PocketBase, Ollama, Piper / Edge TTS, SearXNG, Nginx, and optional Flux image generation when GPU hardware is available.

🔒 School-Controlled Infrastructure

When configured for local providers, AI and data services can run on school-controlled hardware. FERPA posture still depends on district policy, contracts, retention, and configuration.

🔌 Pluggable Backends

Choose your AI provider: Gemini, Ollama, OpenAI, Claude, or any OpenAI-compatible endpoint. Switch backends in one click — no code changes for supported provider configurations.

Varies Hardware budget

Quote-based Existing subscription stack

Equity First — Local-First Architecture

Slide 1 of 22

Universal Design for Learning

400+ documented capabilities organized by UDL Principle. See all features →

Adventure Mode

Turns reading into an interactive RPG with 3D dice, storybook export, and inventory tracking.

Word Sounds Studio

Complete phonemic awareness suite: Segmentation, Blending, Isolation, Rhyming, and Sound Mapping.

Exit Tickets & Quizzes

Generate MCQs with DOK tuning and export directly to Canvas (QTI format).

Interactive Games

Review Game, Bingo, Memory Match, Crossword, Word Search, and Concept Sort.

Leveled Text

Rewrites text for K-12 levels with Immersive Reader, Dyslexia Fonts, and Grammar Highlights.

Escape Room

Teams solve 6 puzzle types to escape a themed room with clue chains and streak bonuses.

Gemini Bridge

Auto-generates sandboxed coding environments with Python, React, and p5.js — directly from lesson content.

Democracy Mode

Students vote on story paths, choices, and directions — with live group differentiation that pushes different content to different groups in real time.

Oral Fluency Coach

Students read aloud, AI calculates WCPM and shows a color-coded accuracy heat map for RTI monitoring.

STEM Lab

141 STEM plugin files / 142 registered STEM IDs across math, science, coding, music, economics, CTE, life skills, and applied simulations — browser-native and Desktop-ready.

Modular Architecture

Hub-and-spoke framework lets developers build independent modules that inherit Gemini AI, translation, and data storage — zero backend required.

63 Language-Pack Files

UI keys are mirrored across the current language pack files, with bilingual views and AI-assisted content translation where configured.

Immersive Reader

Dyslexia fonts, line focus masks, syllable splitting, and grammar color-coding, with WCAG-oriented design and published audit/VPAT evidence.

Exit Tickets

AI-generated DOK-leveled quizzes with instant grading and one-click Canvas QTI export — a paperless workflow with less manual data entry.

Teacher Dashboard

Class analytics, RTI tier tracking, and learner profiles support local-first storage paths and exportable evidence for school review.

BehaviorLens Clinical Suite

specialized clinical workflows — IOA calculators (5 methods), DTT data sheets, preference assessments, scatterplot analysis, effect size calculators, IEP goal generators, MTSS tier management, and AI-powered observation coaching.

Report Writer

Psychoeducational report wizard with Fact-Chunk accuracy pipeline — extracts scores, scrubs PII, and generates clinically-accurate narrative sections with AI audit verification. Multi-language export included.

Research Suite

Built-in Research Mode with study setup (IRB tracking), auto-triggered student/teacher/parent surveys, CBM score import, fidelity logging, math and reading fluency probes, and one-click CSV research data export.

AlloStudio

Create born-accessible visual documents with real text, reading-order controls, alt/decorative image states, and transparent process history.

Cinematic Studio

Turn source documents into grounded storyboards and captioned, translated videos rendered in the browser.

Open Groove Studio

Compose classroom-safe beats and music in a browser groovebox with synthesis, samples, and license-aware workflows.

Professional Development

Publish and complete community-authored PD modules with AI co-authoring, progress tracking, and certificates.

View 400+ Documented Capabilities

Built by an Educator, for Educators

AlloFlow was designed by a school psychologist who understands the realities of the classroom.

Free Consultation

Want help setting up AlloFlow for your classroom or district? I offer free 1-on-1 consultations to walk you through the platform, answer questions, and help you get started.

Schedule a Session

Share Your Story

Had a positive experience using AlloFlow in your classroom or at home? I would love to hear how it has impacted your learners!

Contact Aaron

Simple, Transparent Access

AlloFlow has no software license fee and is open source. The web Gemini Canvas path may require an eligible Google account; Desktop and local-provider paths can run without one.

Keyless web path

Gemini Canvas

No AlloFlow license fee. Google controls account eligibility, availability, quotas, and plan terms; school accounts may require administrator enablement.

Recommended local-first path

AlloFlow Desktop

No AlloFlow license fee. Local models avoid cloud API charges; hardware, electricity, support, and any optional cloud provider remain your responsibility.

School-controlled deployment

Hosted or School Box

No AlloFlow license fee. Budget separately for hosting or hardware, security operations, support, and usage-based AI services.

Compare Deployment Paths

Gemini availability and limits can change. Verify current details in the official Gemini limits guide and Google AI plan page.

Disclaimer: AlloFlow is an independent open-source project and is not affiliated with, endorsed by, or sponsored by Google.

Data Privacy & Architecture

AlloFlow is designed with security in mind, putting control in the hands of educators.

File-Based Persistence

Because Gemini Canvas environments don't persist internal databases across sessions, AlloFlow uses a "sneakernet" architecture. You save and load your activities directly as standard JSON files to your hard drive. This makes sharing custom lessons with other teachers or deploying activities to students as simple as sharing a file.

Google Education Infrastructure

By running as an artifact within Gemini Canvas, AlloFlow piggybacks entirely on your district's existing Google Workspace for Education security agreements. API calls happen directly through Google's infrastructure; privacy obligations still depend on your district's Google Workspace configuration, contracts, retention settings, and actual use.

Local-First Student Data

Student progress, XP, quiz scores, and session data can stay on-device in local-first paths. Cloud, hosted-session, and provider-backed paths vary by deployment. Teachers can export data as JSON files for record-keeping. The Student Portal uses generated codenames instead of real names, keeping learner identities private by default.

Ready to Transform Your Classroom?

Join a growing community of educators saving time and reaching every learner.

Star on GitHub