New Delhi: RailTel Corporation of India Ltd has secured a ₹27.02 crore contract to build a statewide legal AI assistant platform for the Andhra Pradesh government, bringing the railway PSU into the state’s efforts to deploy artificial intelligence across public administration.
The project, awarded by Andhra Pradesh Technology Services Ltd (APTS), will be hosted on Google Cloud Platform and cover the development, implementation, hosting and managed services of the platform.
RailTel, a Navratna public sector undertaking under the Ministry of Railways, received the Letter of Acceptance (LoA) on November 21, 2026.
Valued at ₹27,02,56,350, including taxes, the domestic contract is scheduled for execution by May 9, 2029.
RailTel has also clarified that the awarding entity has no promoter or promoter-group interest and that the contract is not a related-party transaction.
The project is expected to create a common AI-based legal information and assistance system for state government departments, public sector undertakings, universities, commissions and other autonomous institutions.
Andhra Pradesh Plans AI System Across Government Legal Departments
The scale of the proposed legal AI assistant platform is reflected in the volume of records it is expected to process.
According to the tender documents published on the Andhra Pradesh eProcurement portal (Tender ID 985766), the project involves digitising, ingesting and indexing approximately 7.47 terabytes of existing legal records.
The system will also draw information continuously from multiple judicial and government platforms, including the Online Litigation Management System (OLCMS), e-Courts, the National Judicial Data Grid (NJDG), the Supreme Court, tribunals and departmental systems.
Rather than functioning as a standalone chatbot, the proposed platform is designed to support several stages of government legal work. Its seven modules cover data integration, a grounded legal knowledge base, conversational AI, litigation lifecycle management, contempt prevention, executive analytics, AI governance and integration with existing systems.
Government legal teams will be able to use the assistant in English and Telugu for preparing case briefs, tracking case chronologies, conducting legal research, compiling hearing packs and assisting with document drafting.
The tender also specifies citation grounding and human-in-the-loop review, allowing legal professionals to verify AI-generated information and drafts.
Google Cloud Deployment With Data Localisation Requirements
The Andhra Pradesh government has specified that the platform will be deployed within its Google Cloud tenancy. Mumbai will serve as the primary cloud region, while Delhi NCR will be used for disaster recovery.
The tender requires project data to remain within India.
The implementation period is structured around six months for statewide deployment, followed by 24 months of hosting and managed services.
The government has retained the option to extend the managed services period by up to another 24 months.
The platform is also required to be model-agnostic, allowing the government to avoid dependence on a single AI model for its legal assistance requirements.
Why the Tender Had Stringent AI and Legal Experience Requirements
APTS issued the Google Cloud tender on September 4, 2026, on behalf of the state’s Real Time Governance Society. Bids closed on October 15 after a pre-bid meeting and multiple corrigenda. The tender carried an earnest money deposit requirement of ₹25 lakh.
The eligibility criteria sought companies with established IT operations and experience in deploying generative AI applications.
Bidders were required to be Indian companies or limited liability partnerships with an average annual IT turnover of at least ₹100 crore over the preceding three financial years.
The tender additionally required authorised Google Cloud partner status, backed by project-specific authorisation from Google Cloud India.
On the project experience front, applicants needed at least two production generative AI or large language model projects in India, each valued at ₹3 crore or more, during the preceding four years.
They also needed at least one legal, judicial or litigation-domain project worth ₹1 crore or more in the preceding five years, along with a committed full-time onsite legal domain expert.
These conditions placed emphasis on both enterprise-scale AI deployment capabilities and familiarity with legal-sector workflows.
A Parallel Azure Tender Was Also Floated
Andhra Pradesh’s legal AI procurement was not limited to Google Cloud. On September 4, APTS also issued a near-identical tender for the same statewide platform on Microsoft Azure.
The Azure tender, listed under Tender ID 985771, followed the same bidding schedule and carried an identical ₹25 lakh EMD requirement.
However, RailTel’s LoA relates specifically to the Google Cloud assignment.
RailTel’s Project Business Accounts for Nearly Two-Thirds of Revenue
For the quarter ended September 30, 2026, the company reported standalone revenue from operations of ₹1,146.14 crore, up 20.5% from ₹951.36 crore in the corresponding quarter last year.
Net profit, however, declined 9.9% year-on-year to ₹68.56 crore. The quarter included an exceptional loss of ₹32.17 crore.
RailTel’s Project Work Services segment generated ₹749.20 crore, accounting for approximately 65% of quarterly revenue. The segment’s revenue stood at ₹583.83 crore in the year-ago quarter.
Telecom Services contributed ₹396.94 crore, compared with ₹367.53 crore previously.
For the first half of FY27, RailTel reported revenue from operations of ₹2,039.41 crore.







