NHAI turns to AI and open source to preserve institutional knowledge lost to officer transfers

Account subscription benefits alongside Premium Stories, Editorials, Opinions and more.

Unlock these with Subscription

The View From India Looking at World Affairs from the Indian perspective.

First Day First Show News and reviews from the world of cinema and streaming.

Today's Cache Your download of the top 5 technology stories of the day.

Science For All The weekly newsletter from science writers takes the jargon out of science and puts the fun in!

Data Point Decoding the headlines with facts, figures, and numbers

THEdge At the cutting edge of education and careers

Health Matters Ramya Kannan writes to you on getting to good health, and staying there

Gender Agenda Stories from beyond the binary.

The Hindu On Books Books of the week, reviews, excerpts, new titles and features.

Updated - October 21, 2026 03:45 pm IST - Chennai

‘Marg Sarthi’, NHAI’s internal AI assistant, allows officers to query project data and policy documents through a conversational interface.

File photo for representational purposes only. | Photo Credit: The Hindu

The National Highways Authority of India (NHAI) is using artificial intelligence (AI) to help officials navigate the organisation’s policies, contracts and highway project data, while working to preserve institutional knowledge that can be lost when officers are transferred.

Mayank Kumawat, an AI engineer at NHAI, and Deep Saikia, Deputy Manager (Information Technology) at the authority, outlined the initiative at the recent IndiaFOSS 2026, the annual conference organised by FOSS United.

Their presentation described how NHAI’s internal AI team is using open-source software to build tools for everyday administrative and technical work, and how it is exploring ways to make employees’ knowledge reusable across the organisation.

Mr.

Kumawat is part of NHAI’s AI Cell, where he helps build internal AI tools, including the authority’s assistant ‘Marg Sarthi’.

According to his IndiaFOSS, he has worked on applied AI, education technology and public-sector systems, was a lead contributor to the Ministry of Statistics and Programme Implementation’s MCP initiative, and built an adaptive AI learning system called SMART in 2023 in collaboration with XLRI and Harvard faculty.

Deep Saikia (left), Deputy Manager (Information Technology), and Mayank Kumawat, an AI engineer at NHAI. | Photo Credit: Special arrangement

Mr.

Saikia, a Computer Science and Engineering graduate from the National Institute of Technology (nit), Silchar, works on AI-based tools, digital transformation and workflow automation at NHAI.

His work focuses on applying technology to public infrastructure and large-scale institutional processes.

Mr.

Kumawat said that when he joined NHAI, he found that officials had to consult multiple sources to carry out routine tasks.

These included laws, Ministry of Road Transport and Highways circulars, more than 1,800 internal NHAI policy circulars, and standards issued by the Indian Roads Congress.

More complex tasks, such as approving contractor bills, drafting notices, handling land acquisition and obtaining environmental clearances, required officials to navigate several rules and documents.

He said officials often maintained their own folders of relevant circulars and guidelines, updating them as new documents were issued.

But frequent transfers and deputations meant that incoming officers had to rebuild this knowledge base.

Mr.

Saikia said the problem was not a lack of expertise within NHAI, but the difficulty of making that expertise reusable across the organisation.

“The problem is that the knowledge is not always reusable,” he said during the presentation, adding that when officers move or are transferred, some of the context they have accumulated moves with them.

The team is experimenting with converting recurring administrative and technical processes into reusable workflows.

Mr.

Saikia said experienced officers could explain how particular processes work, identify the documents and approvals required, and have other experts review those procedures before they are incorporated into the system.

He said this could allow a workflow understood in one office to be used in another, and help prevent mistakes identified by one officer from being repeated elsewhere.

Mr.

Kumawat demonstrated how ‘Marg Sarthi’, NHAI’s internal AI assistant, allows officers to query project data and policy documents through a conversational interface.

In one demonstration, Mr.