Arvind Pani
Vision & StrategySets the company's direction and long-term bets.
Company
We've built Indian-language AI before. Now we're building the data behind it.
Ikyam is a data intelligence company. We source, verify and certify the training data that AI models need in the languages and voices most of the world actually uses.
Our story
Reverie, Jio, and a gap that never closed.
Our founding team built Reverie Language Technologies, one of India's earliest language-AI companies, later acquired by Jio. Between them, they've spent over 100 years building AI for Indian languages, and shipped it at national scale. One co-founder has spent 30 years in Indian-language computing alone.
That work taught us where every model breaks: not on compute, but on data. Clean, consented, model-ready data in the world's under-served languages barely exists. Ikyam exists to build it, properly, from day one.
Why this matters
Good AI needs data most companies can't get.
Every frontier lab and enterprise building AI for the next billion users hits the same wall: the languages, accents and contexts they need aren't reliably represented anywhere.
We built Ikyam to be the vendor that solves this properly: ethical sourcing, a technology platform that engineers in quality, and a vetted community of native-language experts, working as one system.
The team.
Nine people covering sourcing, technology, community and go-to-market, all with a shared history building Indian-language AI.
Sai Nanduri
Revenue & BusinessLeads go-to-market strategy and growth.
Thomas Hyland
Big Tech PartnershipsBuilds relationships with frontier labs and big tech.
Prantik Nayak
Delivery HeadOwns delivery outcomes across people, processes, timelines, and quality.
Anurag Behera
CommunityRuns the network of vetted annotation experts.
Vivekananda Pani
Technology & MentorshipGuides the technical roadmap and engineering culture.
Pranjal Nayak
Data SourcingOwns raw data ingestion, sourcing strategy and consent.
Bhupen
PlatformBuilds the tech-enabled processing pipeline.
Chintan Parikh
EnterpriseLeads enterprise accounts and delivery.
Where we are today
Built. Running. Ready to scale.
Hours of consented, compliant data collected
Vetted native-language annotators onboarded
Platform features purpose-built for annotation
Years of combined language-AI experience
Data is sourced with consent from day one, and stored and processed in India in line with the DPDP Act.