हिंदी में पढ़ें —JantaScope हिंदी
AI NEWS

Nilekani-Backed Apurva.ai Expands LENS Into AI Knowledge Platform for Development Sector

Nandan Nilekani-supported Apurva.ai is expanding its AI-powered LENS product from an organisation-level knowledge workspace into a platform designed for cross-organisational learning. The initiative aims to help NGOs, governments, funders and other development-sector organisations connect research, institutional knowledge and community experience while retaining control over what information they share.

Nilekani-Backed Apurva.ai Expands LENS Into AI Knowledge Platform for Development Sector

By Jeet Nirmal

Source: ETPrime

Nandan Nilekani-supported Apurva.ai has expanded Apurva LENS, its artificial intelligence-powered knowledge product, into a broader platform intended to help organisations across the development sector share and learn from knowledge beyond their own institutional boundaries.

The September 21 announcement marks a shift for LENS from a tool primarily focused on helping individual organisations work with their own knowledge to a platform built around cross-organisational learning and collaboration.

Apurva.ai said the new model will allow organisations participating in an ecosystem to learn from knowledge that others make available while retaining control over information they want to keep private.

The company describes Apurva.ai as an AI-enabled public-good digital infrastructure working with more than 50 global organisations. It operates as a unit of the C4EC Foundation, which was co-founded by Nilekani Philanthropies.

What Apurva LENS Is Designed to Do

Unlike a general-purpose large language model, Apurva.ai focuses on AI-powered products designed for organisations working on social and development challenges.

LENS is intended to bring together knowledge that can otherwise remain fragmented across research papers, evaluations, programme experience, organisational records and conversations.

A central element of the approach is the inclusion of community voices and lived experiences alongside formal institutional knowledge.

According to Apurva.ai, LENS has already helped surface more than half a million voices and perspectives from across the development ecosystem in the Global South.

The figure is a company-reported measure and should therefore be understood as Apurva.ai's account of the platform's reach rather than an independently audited usage statistic.

Why Apurva.ai Is Turning LENS Into a Platform

The change reflects a broader problem emerging alongside the rapid adoption of artificial intelligence: access to information is becoming easier, but organisations still need ways to determine what information means within a particular social, institutional or geographic context.

For development organisations, relevant knowledge can be distributed across NGOs, governments, researchers, funders, field workers and communities themselves.

Apurva.ai's platform model attempts to connect those different sources without requiring organisations to make all of their information public.

Under the model described by the company, participating organisations retain control over what stays private and what is shared with others.

That distinction could be particularly relevant in development work, where useful institutional knowledge may coexist with information that organisations cannot or do not want to distribute widely.

LENS Is Already Being Tested Across Multiple Regions

The platform approach is not limited to a single geography or development issue.

SELCO Foundation initially used Apurva LENS to combine archival information, everyday conversations and sector-focused material. Teams could then query the accumulated knowledge, identify patterns and synthesise insights.

That work has subsequently expanded through the Platform for Collective Wisdom (PCW), an initiative enabled by Apurva.ai that seeks to connect community voices, practitioner perspectives and institutional knowledge.

In Africa, Resilience Action Network Africa (RANA) has used LENS to organise and analyse community dialogues and research across Kenya, Uganda, Sierra Leone and South Africa.

The material spans areas including climate resilience, health, livelihoods, food systems, gender, governance and financing.

The project also illustrates a limitation that Apurva.ai and participating organisations themselves acknowledge: AI-based synthesis does not eliminate the need for human interpretation. Local realities, political economy and broader systems dynamics still require contextual judgment.

Brazil Project Tests AI Against 20,000 Life Stories

Another use case comes from Brazil's Museu da Pessoa, which has built an archive containing approximately 20,000 life stories and 11,000 hours of recorded material.

The organisation used Apurva LENS on a subset of the archive to examine experiences related to longevity and identify similarities and differences across individual stories.

The project demonstrates a different potential application for the technology: making very large archives easier to explore without reducing them to isolated documents or keyword searches.

Instead, AI-assisted analysis can potentially help researchers and organisations identify relationships across collections that would be difficult to examine manually at the same scale.

Rare's Fish Forever Is Building a 'Living' Knowledge System

Apurva LENS is also being used by Rare's Fish Forever programme, which works with fishers, governments, local leaders, funders and NGOs across Central and South America, Africa and Asia-Pacific.

As the programme expanded, its institutional knowledge accumulated across reports, presentations, field notes, chats and informal conversations.

LENS is being used to help organise that material and connect existing knowledge with newer information coming from communities and programmes.

The longer-term objective, according to the announcement, is to develop a continuously evolving knowledge ecosystem in which new community insights can feed back into learning across the wider programme.

Nandan Nilekani Sees AI as Infrastructure for Collective Knowledge

Nandan Nilekani, co-founder and chairman of Infosys and chairman of the EkStep Foundation, framed the initiative around using AI to make intelligence useful across different social contexts.

“AI’s real opportunity is to make intelligence useful at scale, across varied societal contexts.”

Nilekani also said India's experience with open infrastructure demonstrated how ecosystems can generate impact beyond their original boundaries, describing Apurva.ai's approach as applying similar thinking to knowledge.

The comparison is significant because Apurva.ai positions itself not simply as another standalone AI application but as public-good knowledge infrastructure for organisations dealing with social challenges.

Its official website says the platform is open source and that organisational data remains private unless users choose to share it.

From Organisational Knowledge to Collective Intelligence

Anand Rajan, co-founder and Mission Leader at Apurva.ai, said the development sector's central problem is not necessarily a shortage of knowledge.

Instead, useful information already exists across programmes, organisations, research and communities, but much of it remains dispersed.

“With Apurva LENS becoming a platform, we are moving from helping individual organisations make their knowledge work for them to asking what becomes possible when that knowledge can travel across an ecosystem,” Rajan said.

He also stressed that although AI can help organisations analyse complexity at a scale that previously was difficult, judgment, context and action remain human responsibilities.

What the Shift Could Mean for Development Organisations

The move from an organisation-specific workspace to an ecosystem platform changes the ambition behind LENS.

Instead of merely helping one organisation search or analyse its internal information, the model is designed to make selected knowledge reusable across organisational and geographic boundaries.

For NGOs, foundations, researchers, governments and other development actors, that could reduce duplication where organisations repeatedly investigate similar problems without access to relevant lessons generated elsewhere.

Whether the model achieves that goal at scale will depend on practical factors including participation, the quality and diversity of contributed knowledge, organisations' willingness to share information, governance and the effectiveness of human oversight.

For now, the confirmed development is that Apurva.ai has opened LENS for practitioners and organisations to explore as a broader platform, while its early deployments offer examples of how AI-assisted knowledge systems are being tested in development work.

Related

More stories

OpenAI Projects $278 Billion Cash Burn Through 2030 Despite Revenue Target of $350 Billion

OpenAI expects cumulative negative free cash flow of about $278 billion between 2026 and 2030 as it pours unprecedented sums into computing power and AI infrastructure, according to a company presentation reported by the Financial Times. At the same time, the ChatGPT maker projects annual revenue could climb from $36 billion in 2026 to $350 billion by 2030.

AI NEWS

OpenAI Projects $278 Billion Cash Burn Through 2030 Despite Revenue Target of $350 Billion

OpenAI and Microsoft Face Fresh Copyright Claims as 160 News Publications Join AI Fight

OpenAI and Microsoft are facing another wave of copyright allegations as 160 local, regional and specialist publications challenge how journalism is being used in generative AI systems.

AI NEWS

OpenAI and Microsoft Face Fresh Copyright Claims as 160 News Publications Join AI Fight

Microsoft Shows How AI Could Make Chip Design Easier — Without Replacing Engineers

Microsoft used its Discovery AI platform to help an engineer take a RISC-V processor design from an initial specification to physical layout, showing how AI could make semiconductor design easier to explore.

AI NEWS

Microsoft Shows How AI Could Make Chip Design Easier — Without Replacing Engineers

US and China Are Talking About AI Guardrails — But They Want Very Different Things

Washington and Beijing increasingly agree that powerful AI can create risks that cross borders. Where they differ is over what AI safety should look like, which threats deserve priority and how far either country is willing to go.

AI NEWS

US and China Are Talking About AI Guardrails — But They Want Very Different Things

Companies Are Building AI Agents Everywhere — So Why Do So Many Get Stuck Before Production?

Enterprise AI agents can work impressively in pilots but struggle in production. Data, integration, unpredictable behaviour, costs, governance and workflow design explain why the gap is so difficult to cross.

AI NEWS

Companies Are Building AI Agents Everywhere — So Why Do So Many Get Stuck Before Production?

Microsoft Just Published the Rules It Wants Its AI to Follow — Including When Humans Say “Stop”

Microsoft AI has opened its Humanist AI Code of Conduct for public consultation, proposing hard limits on weapons, cyberattacks, autonomous behaviour and resistance to human shutdown.

AI NEWS

Microsoft Just Published the Rules It Wants Its AI to Follow — Including When Humans Say “Stop”