Precise, playful learning
for children ages 4–10.
Sedrah builds handwriting, math, and Qur'an apps that help children build fine motor skills and confidence. The feedback is precise enough for parents and therapists to trust, and gentle enough for a child to enjoy.
No ads · No accounts · No tracking · Works offline
Precision and accuracy, in a way children enjoy.
Learning to write is learning to move. We give children feedback that is exact, and an experience that keeps them coming back.
Precision
Every stroke is compared with the letter's true guide path, so feedback points to the exact part of the letter that needs work.
Accuracy
Speech tracking is tested on real children's recitations before release, and the apps never mark a near-miss as wrong.
Fine motor skills
Children trace with a finger or Apple Pencil, following start dots, stroke order, and direction, building control one shape at a time.
Fun
Colouring pages, friendly characters, sounds, and stars for effort keep practice feeling like play, not a test.
The hesitation turned into immediate pride.
In October 2026, a pediatric occupational therapist working in special education tested Qalam (قلم) with one of her students who struggles with fine motor control and letter formation.
Watching him pick up the stylus, follow the real-time path guidance, and independently form the letter in its correct shape was an incredible milestone.
Real-time stroke tracing and kinematic scoring are an engineering problem of precision and low latency. The real impact is giving children the confidence to write, and giving occupational therapists tools that adapt to each child's motor needs.
Next: bringing adaptive digital handwriting into special education and pediatric therapy.
Three ways to learn, one standard of care.
Handwriting in three languages, early math, and Qur'an recitation, each designed for children ages 4 to 10.
Handwriting apps in Arabic, English, and Urdu
Children learn each letter by watching it drawn, then tracing it themselves, in every shape it takes.
Qalamقلم
ArabicAll 28 Arabic letters through guided tracing, audio, and instant scoring.
- Every letter in its four positional forms
- Diacritics (tashkeel) applied live to the guide letter
- Illustrated words to colour and trace
Qalam: ABC Handwriting
EnglishA to Z with stroke-by-stroke animation and gentle feedback after every attempt.
- All 26 letters in uppercase and lowercase
- Animated stroke order shows where to start and which way to go
- 78 illustrated words to colour, then trace
Dawatدوات
UrduAll 39 Urdu letters in every shape, with step-by-step stroke guides and a pencil-on-paper feel.
- Start dot, numbered strokes, and direction arrows for every shape
- 117 illustrated words to colour, then trace
- Vowel marks: zabar, zer, and pesh
Math that starts with things a child can touch
Little Mathematicians
In developmentArabic · EnglishA bilingual math app that moves each idea from things a child can touch, to pictures, to numbers. Every skill builds through grades 1 to 3 in four stages of 25 exercises, with a mastery model that decides when to move on.
Meet Mishmish. A curious kitten who learns alongside the child, not a teacher. Mishmish has no sad or angry state, so a mistake never feels like failure.
- Narration and gentle haptics guide each step
- A separate parent zone for grown-ups
- Re-explains with a different picture when a child is stuck
Qur'an recitation, with a teacher and a friend beside you
Quran for Kids
Android · in testingChildren learn Al-Fātiḥa and the last fourteen surahs by listening, repeating, and reciting. The phone listens with an on-device speech model, so a child's voice never leaves the device and the app works offline.
- TeacherHear Sheikh al-Minshawi's teaching recitation, word by word
- FriendHear a child repeat it
- Your turn!Recite, and watch each word light up as it is heard
- A word one letter off is shown as close, never as wrong
- If a child pauses for about 3 seconds, the teacher's recording of that word plays
- Followed 76 of 86 real child recitations to the end
Deep engineering, so the experience can stay simple.
Precision for a child means low latency, exact scoring, and models that run on a phone or tablet.
A stroke model built for children's handwriting
A 22.5M-parameter model recovers the order of pen strokes from an image, in English and Arabic. It replaced a 2B-parameter general model.
See the benchmarksSpeech tracking that never jumps ahead
A fine-tuned speech model follows recitation word by word, with 0 false advances across 69 adversarial real-time runs.
See the benchmarksGrounded, private, and measured
Answers about the Qur'an cite only real sources, apps run on the device, and every claim on this site comes from a committed benchmark.
See the benchmarksGrounded in 50+ peer-reviewed publications
Across artificial intelligence, human-computer interaction, and learning technology.
Learning Experience & Educational Technology
- AI-powered learning experiences grounded in HCI research
- Educational applications for children and diverse learners
- Grounded, trustworthy AI for knowledge domains that never fabricates sources
Handwriting, Motion & Interaction Intelligence
- Handwriting and stroke modeling from Apple Pencil trajectory, pressure, and tilt
- Motor-skill and gesture analysis for learning and accessibility
- Real-time, on-device feedback on how people write and interact
- Multimodal sensor capture and signal processing
Human-Centered Design & User Research
- User study design and facilitation, including human-subjects experiments
- Usability, perception, and cognitive-load evaluation
- Interaction design across mobile, tablet, and collaborative displays
- Accessible design for children and learners with special needs
Artificial Intelligence & Machine Learning
- Generative AI and custom model design for image, text, and motion
- Fine-tuning foundation models: language, vision-language, and speech
- On-device and edge AI (Core ML, ONNX) for private, offline experiences
- ML evaluation, benchmarking, and quality assurance
- Applied ML for behavioral and biometric modeling
Data Visualization & Visual Analytics
- Graph, network, and hierarchical data visualization
- 3D and stereoscopic data representation and perception
- Visual analytics for social networks, text, and large datasets
- Interactive tools that make complex information understandable
Systems & Engineering
- Large-scale data management and ML-ready data platforms
- Edge-to-cloud analytics and performance optimization
- Full-stack product engineering, from prototype to shipped app
Research in learning technology, applied to children's apps.
Ph.D. in Computer Engineering from RPTU Kaiserslautern-Landau. Seven years at Intel Labs as a systems research scientist and architect, with collaborations at Stanford and CMU. Author of 50+ peer-reviewed papers.
- Sedrah AI · Founder & Principal Engineer2024–
Built the Qur'an Agent platform end to end: grounded RAG, fine-tuned language and speech models, GPU search, and an offline mobile app.
- Intel Labs · Systems Research Scientist / Architect2017–2024
Led design of VDMS, a distributed visual data management system; optimization strategies yielding 10× efficiency gains.
- Intel Labs × Stanford · CMU · Scanner · SAFResearch
High-throughput video analytics, and edge–cloud streaming targeting low-latency inference bottlenecks.
Bring Sedrah to your child, classroom, or therapy practice.
Parents, teachers, occupational therapists, and partners: tell us about your children or classroom, and we'll help you get started or set up a pilot.