\n \n

\n\n# KODAAI

Voice-First AI for Accessible Education, Industrial Voice Command, and Turkish Language Technologies

Türkçe · Project Vision · Roadmap · Accessibility · Website

KODAAI is an independent voice-first artificial intelligence and accessible technology initiative building systems that let people interact with digital services, learning environments, and operational workflows through natural voice.

Our first product, KODAAI Education, is a voice-first learning platform designed with a 100% accessibility target for blind and visually impaired learners preparing for Türkiye’s EKPSS examination.

KODAAI also conducts parallel research and development in:

• industrial voice command and control;
• Turkish large language models;
• Turkish text-to-speech technology;
• accessible and assistive AI;
• applied artificial intelligence for education and industry.

This is KODAAI’s public project and collaboration repository. Proprietary source code, private datasets, production infrastructure, user data, voice recordings, and unpublished research are not included.

\n \n

\n\n## Product and Research Portfolio

Initiative	Focus	Status
KODAAI Education	Voice-first accessible learning, quizzes, and exam preparation	First product · private beta
Industrial Voice Command Center	Voice-driven information access and operational workflows for industrial organizations	Active R&D · not publicly released
Turkish LLM	Turkish command understanding, dialogue, and domain-specific intelligence	Research and development
Turkish TTS Engine	Natural, intelligible, and accessibility-oriented Turkish speech synthesis	Research and development
Applied AI	Human-centered AI for education, accessibility, and industry	Continuous R&D

Why Voice-First?

Most digital products are designed visually and receive voice support later. KODAAI starts from a different premise: voice is the primary interaction layer.

This approach is intended to support:

• independent access for blind and visually impaired users;
• low-friction interaction in hands-busy or eyes-busy environments;
• faster access to operational knowledge;
• Turkish-first language and speech experiences;
• resilient local and offline-capable workflows where appropriate.

First Step: KODAAI Education

KODAAI Education enables learners to navigate lessons, answer topic quizzes, complete timed mock examinations, hear results, and control playback using voice.

Current focus areas include:

• EKPSS preparation for blind and visually impaired candidates;
• voice-controlled reading, quiz, and exam flows;
• deterministic local audio assets;
• web and Android experiences;
• keyboard and screen-reader compatibility;
• user-tested accessibility improvements;
• privacy-aware progress and session telemetry.

Learn more in the Turkish project overview.

Technology Direction

KODAAI explores and combines:

• voice user interfaces and conversational interaction;
• speech recognition and command routing;
• Turkish natural language processing;
• Turkish LLM and TTS research;
• real-time WebSocket systems;
• accessible web and mobile interfaces;
• local-first and offline-capable speech workflows;
• secure AI integration for industrial environments.

Accessibility Principles

1. Voice is a primary interface, not an optional add-on.
2. Critical actions must not depend on visual feedback alone.
3. Keyboard and screen-reader paths remain available.
4. Ambiguous commands require safe recovery or confirmation.
5. Accessibility decisions are validated with target users.
6. Personal data and voice recordings follow data-minimization principles.

Read the full Accessibility Statement.

Collaboration

KODAAI is open to conversations with:

• organizations serving blind and visually impaired people;
• universities and accessibility researchers;
• education and assessment institutions;
• industrial organizations exploring secure voice interfaces;
• Turkish NLP, LLM, TTS, and speech researchers;
• responsible technology partners and investors.

See CONTRIBUTING.md for collaboration routes.

Repository Policy

This public repository is a project showcase and collaboration surface. It does not grant access to KODAAI’s private core implementation.

No open-source license has been granted. Unless explicitly stated otherwise, all content and project materials are protected by copyright and may not be copied, modified, redistributed, or used commercially without written permission.

Discovery Keywords

Voice-first AI, accessible artificial intelligence, industrial voice command center, Turkish LLM, Turkish language model, Turkish TTS, Turkish text-to-speech, speech technology, accessible education, blind accessibility, visually impaired education, assistive technology, EKPSS, accessible EdTech, human-computer interaction, natural language processing, applied AI.

Contact

• Website: kodaai.com.tr
• GitHub: @Drsuspect
• Project lead: Murat GÜNEY LARRANAGA

Rights

Copyright © 2026 Murat GÜNEY LARRANAGA. All rights reserved.

KODAAI, KODAAI Education, related concepts, documentation, product names, and unpublished technologies remain proprietary unless a separate written license states otherwise.