Indian-language speech
Speech recognition and language models tuned for the way people across India really talk — accents, dialects and languages mixed mid-sentence.
Niranjan Technosoft is a Mangaluru software company building AI products across fields. Our first, Veda, turns long videos into captioned short clips — accurate even when a speaker moves between Hindi, Kannada and English in the same sentence.
Veda is for creators, media houses and organisations who record long-form video and need short clips for Reels, Shorts and other vertical feeds.
We build AI products for real problems across many fields. Veda is the first to launch, and more are in development.
Speech recognition and language models tuned for the way people across India really talk — accents, dialects and languages mixed mid-sentence.
Systems that watch and listen to long recordings, find the strongest moments and turn them into ready-to-post clips.
We're developing new products that apply AI to problems in other industries, built with the same care for Indian users and languages. More to share soon.
Durable processing pipelines on AWS that handle long uploads reliably, built with FastAPI, Temporal and PostgreSQL.
Founder
A postgraduate in Business Management from Mumbai University, Nithin has spent over 23 years in planning, strategy and execution as the founder of Niranjan Techno Fab, delivering complex, large-scale installations on tight timelines. He now brings that operational discipline to building software and AI products.
Co-founder, Veda
A student and keen technologist, Aarav co-founded Veda and works on its user experience and AI integrations, bringing a younger generation's view of how people create and watch video.
For partnerships, enterprise use of Veda, or working with us, write to us.
contact@niranjantechnosoft.in