Our Story
Kekar was born from a simple belief — every Indian language carries centuries of wisdom, poetry, and soul. We built a platform where that heritage meets modern AI.
🎯 What We Do
Kekar is an AI-powered storytelling and audio platform built specifically for India. We combine large language models with neural text-to-speech technology to let anyone — student, teacher, grandparent, or creator — generate, listen to, and share stories in their own language.
From a simple topic or voice prompt, our platform writes a vivid cultural story, narrates it in a chosen voice, and even reads uploaded documents aloud — all in seconds.
🗣️ Languages We Support
We celebrate the linguistic diversity of India. Kekar currently supports:
⚙️ Technology
Kekar is powered by state-of-the-art open AI models via the HuggingFace platform. Story generation uses large instruction-tuned language models. Text-to-speech uses Parler-TTS for personalised voices and MMS-TTS for multilingual support. Speech-to-text relies on OpenAI Whisper for accurate transcription across Indian accents.
The platform is built on Laravel, making it fast, secure, and scalable — with a queue-based audio pipeline that handles long documents without keeping you waiting.
🌅 Our Vision
We believe AI should serve regional languages, not replace them. Our goal is to make Kekar the go-to platform for Indian storytelling — a digital daastangoi where oral traditions live on, where grandmothers can hear their own language narrated back to them, and where children discover stories from across India.
In the coming months we plan to expand to more languages, add community story sharing, and bring offline support for low-bandwidth regions.