LankaData has officially released Lanka Data Search, a new public search platform designed to provide fast and intelligent access to over 90,000 public documents and archives in Sri Lanka through the Lanka Data Net (LDN) ecosystem.
The platform enables users to search across a wide range of Sri Lankan public archives, including Court Judgments, Acts and Amendments, Gazettes, Bills, annual and quartely Economic updates released by the Central bank of Sri Lanka and all other official public records through a single unified search interface.
Developed as part of the intelligent data infrastructure of Lanka Data Net, Lanka Data Search introduces contextual and AI-assisted retrieval capabilities aimed at improving public accessibility to Sri Lankan public documents.
In addition to the release of Lanka Data Search, LankaData, in collaboration with its research and technology partners, has also announced the release of Sri Lanka’s first large language trilingual model (LLM) trained in Sinhala, Tamil, and English, including localized language variations such as Singlish and Tanglish.
The model has been designed to better understand Sri Lankan linguistic patterns, mixed-language communication, and local contextual usage that are often not fully supported by global AI systems. According to LankaData, the initiative represents a major step toward building sovereign AI infrastructure and localized artificial intelligence technologies for Sri Lanka.
The newly released language model has been made available as an open-source project through the AI community platform Hugging Face, enabling developers, researchers, and startups to build localized AI applications and tools for Sri Lanka and the broader South Asian region.
Speaking on the launch, LankaData stated that the objective of the initiative is to democratize access to public information while simultaneously building foundational AI technologies tailored to Sri Lankan languages and institutions.
Lanka Data Search is now publicly accessible via:
