Panel Discussion on “Development of Agri Stack: Building blocks of sustained Agriculture Growth”

Amongst several Fintech panels and work sessions at the Global Fintech Festival (GFF) jointly organized by National Payments Corporation of India, the Payments Council of India, and the Fintech Convergence Council, in Mumbai, 5-7th Sept 2023 – a key one was on the potential impact of Agri Data on the Agriculture sector in general and to Agri Fintech in particular.

Link to the session on YouTube: https://www.youtube.com/watch?v=KE7t-es_9tM

The Objective of the Panel Discussion:

The Government of India is developing a National “AgriStack” to help increase farmers' income and improve agricultural efficiency. The Finance Minister of India announced in the budget 2022-23 that a Digital Public Infrastructure (DPI) for Agriculture would be built using open standards to function as an interoperable public good. It is estimated that an Agri Stack as a DPI could provide INR 5-7+ lakh crore benefit for farmers by 2030.

Simultaneously, there has been a lot of work done by several State Governments, Academia, Think Tanks, and the Private Sector in conceptualizing alternate approaches and aspects of an “AgriStack”.

Leveraging the diversity of the eminent panelists, across agriculture, technology, and policy domains, the session objective was to discuss the government’s work on the AgriStack and making it inclusive and usable, keeping in perspective several ongoing use case led efforts around digital public goods in agriculture.

For reference - Digital Public Goods (DPGs) are “atomic” sized building blocks; they are not a solution by themselves but enable end-use applications (Digital Public Infrastructure). DPIs and DPGs can transform the global Agri Food System, addressing critical concerns of climate change, inequity, and access as also the affordability of nutritious food for millions of citizens.

Summary of the Panel Discussion and Next Steps:

I. At the kickoff, expectations from AgriStack were described through several use cases or problem statements by the panelists.

• Farmers have limited information resulting in uninformed decision-making – on crop risks, crop choice based on likely market pricing which is not only based on the previous harvest.
• India Stack changed the financial landscape of our country tremendously with UPI, OCEN, Aadhaar etc. and the expectation from the AgriStack is that it will similarly revolutionise Indian agriculture and pave the way for innovative solutions.
• Help understand the farmer, what s/he grows, and where; combined with weather and soil data, help understand risk mitigation and cultivation potential, yield
• A prime focus of the National AgriStack should be to overcome farmer credit issues and provide innovative solutions to enable ease of access to farm credit.

II. Key Takeaways from the Panel and expectations from the Agri Stack going forward.

a) Agri Data Exchange – a valuable building block for digital agriculture.

An Agri data exchange facilitating the flow of data, between and across private entities and the Government, beyond that provided by the ministry of Agriculture and ensuring strong governance and protection of business interests, will help strengthen open, inclusive public private digital ecosystems, for several use cases.

Ministry of IT (MeitY) is building a Data exchange that can be used for Agriculture, while Telangana recently launched ADeX – a data exchange built by IISc focused on Agri use cases and currently leveraging data sets provided by the GoT. Along with the National Agri Stack multiple data exchanges may mushroom, each with its own niche data sets. The protocols around data sharing, governance and monetisation may evolve and as long as the protocols are “open” and published, the data exchanges will be interoperable.

b) Critical role of the private sector - Agri use cases and monetisation of data:

The work done by the government in building the National AgriStack lays the groundwork for using data and is the starting point for key stakeholders in the Indian agriculture ecosystem to provide digital services that enable better outcomes and results for the farmer.

Use Cases: Private sector has several key roles - to innovate on top of the Agri Stack and widen the range of AI/ML use cases enabled by data - such as access to finance, markets, and advisories to serve smallholder farmers across the country; to consistently layer and leverage the Agri stack through the Data exchange, to ensure access to diverse datasets (like crop futures, warehouse receipts, and other real time farm data). Private entities could become authorised service providers.

Data Sharing: Rich and diverse data sets are needed for complex use cases and while the Government will provide foundational data through its registries, the private sector has to be motivated to consistently contribute data and insights on the data exchange through strong governance frameworks, transparency, and monetisation opportunities for their data.

The future will see private entities that map weather patterns, soil health, as well as seed retailers, pesticide and fertilizer manufacturers, and food processing companies monetise data accumulated by them through a data exchange. Increase in data will facilitate the creation of more digital services and ultimately digital public goods to address challenges faced by the agriculture sector.

DPGs: Private sector should play the lead in the development of “building blocks” or DPGs that are scalable and useful across several Agri use cases, by leveraging the data from the Agri Stack.

It is imperative that the data available with the government, should be available to the private sector for open innovation, through the AgriStack and It was emphasised that the AgriStack would “be a big zero” if the private sector did not leverage it actively and contribute with the vast amount of data and insight it possesses.

c) Integrating the States into the Agri Stack

Agriculture is a state subject; hence, it is crucial to encourage all states to integrate their data and leverage the National AgriStack to provide a uniform national view. Large-scale policy frameworks, and/or a Agri data stack cannot be created in a vacuum, and hence good two way communication and a deep level of trust must exist between the Centre and the States. The roll-out of the National AgriStack is happening in stages, starting with 1 district in each of the 12 chosen states, before complete consumption across the 12 states. The use cases in these districts and the 12 states will pave the way for the integration of other states across the country, thereby facilitating the rollout of the National AgriStack in the balance states of the country.

As different types of data become available, confidence on digital agriculture will grow, and once its benefits and the solutions it can offer to farmers become clearer, there will be an increased buy-in and integration of states across the country with the National AgriStack. It is a “chicken and egg” issue – if the key States and private sector comes on board, it enhances the credibility of the National Agri Stack – making it even more useful for newer use cases through diverse datasets -- which will then in turn attract other States and private users to leverage it.

Given the importance of Land records – a lot of discussion happened on the work done in the last 25 years or so to link the ownership records to land parcels up to 80-85% accuracy levels - the States have a very important role in updating the mutation of land sale records.

Policy frameworks are needed to accelerate the process of data access and sharing. Telangana, for example, has created an agri-data management framework, which will likely become a policy soon, details the SOPs of data sharing, and mechanism for consent management. It was agreed that this policy if useful, should be propagated across other states and avoid duplication.

d) Tenanted Farmers – a challenging techno legal problem.

This is acknowledged as a huge problem. It is estimated that in many states 50% of the land is managed by “informal tenants” and not the farm owners but unfortunately the laws in most states disallow tenancy and hence there is no formal recognition or registration of tenants. Given that official land records state only the landowner, the tenanted farmers slip between the cracks and don’t receive the benefits of Government schemes.

This is a “techno legal” problem, just applying data and technology is insufficient, in the absence of legal overhaul. While states like Andhra Pradesh and Telangana are now beginning to encourage tenancy and developing cohesive tenancy records, the challenges faced by tenanted farmers can only be addressed once concrete legal frameworks for recording land tenancy are developed across states.

Until then, given the vast extent of tenant farmers, the National Agri Stack’s effectiveness to solve the challenges faced by farmers is limited at its best.

e) Consent is vital.

There is a multilayered consent mechanism in place for the farmer data depending on who is maintaining (data fiduciary) and how is maintaining the data and who is using the data. Accordingly, with the development of the data exchange, there will be a need for consent management, clear SOPs on data sharing, and grievance redressal.

III. Discussion on the National AgriStack and its beneficiaries – led by Rajeev Chawla.

a) Very simply, the national AgriStack, similar to the India Stack is a moniker for the collection of basic data sets and tools to manage and share it appropriately, through a set of open APIs. It is the digital foundation being set up by the government to make to improve agriculture in India and enable better outcomes for the farmers. In two years, it will provide the same impact that Aadhar provided by enabling digitalization in the agri sector.

b) India has roughly 170-180 million land owning farmers with approx. 850 million parcels of land (each farmer owns on average 5 parcels) and approx. 700-800 million Indians dependent on agriculture. If we include landless or tenanted farmers or those who work on someone else’s land the numbers dependent on Agriculture swells to 1000 million Indians.

c) The AgriStack will consist of a database of farmers detailing their personal information, their land, location of plots, the crop(s) grown, soil health, weather patterns and other benefits that can be availed by such farmers. Further, it will comprise of a farmer registry and other supporting registries such as crop registry, seed dealers, retailers and wholesalers of fertilizers, importers, pesticide dealers, banks, common service centres and many more. Information especially pertaining to land ownership will be accurate and the farmer registry will be updated by the underlying system of land records and mutation processes.

d) In the case of such landed farmers, the AgriStack will aggregate farmer IDs along with other granular data such as the geographic location of the land (through the 650,000 existing physical maps), the plot size, the kind of crop grown (from the 341 different crops vernacular names), the kind of fertilizers and pesticides used – and the benefits drawn from the Govt. all through a Unified Farmers Service Interface (UFSI). There are 24 data registries planned (see pic below).

e) The data aggregated would help several use cases – an example being ease of access to credit where, presently, farmers need to go to banks physically, and spends 10-15,000 rupees and 30-40 days to convince the bank of the authenticity of their credentials – including land and crop records. Then repeats the process at the sub registrar’s office to register the loan. Jan Samarth

built by the Dept of Revenue, GOI is an example of a “Kisan Credit Card switch” that will make processing loans easier for farmers using a digital portal accessing all relevant data on the Farmer / Farmland / Crop from the Agri Stack on a single platform that connects various banks. The process can be driven electronically through the use of Aadhaar, a farmer ID, satellite data, geo

tagging the location of their plot of land and other aggregated data such as the kind of crop grown, and the farmer’s credit history.

f) Further, through the AgriStack, loan usage will be monitored, including physical verifications thrice a year to ascertain actual crops grown on each plot of land across the Kharif, Rabi and Summer seasons. This will be done by 650,000 people deployed to cover the 850 million farm parcels. Jan Samarth will allow lenders to use their customised front ends to create differentiated user experience much like how payment apps - PhonePe, Google Pay, etc. each have their own user interfaces, but leverage UPI at the backend.

g) Jan Samarth is operational in Karnataka across 5 banks through the use of the data aggregated, would offer loans to individual farmers in real-time, Farmers who were once at the behest of the whims of banks would now have the opportunity to choose between offers from multiple banks. Loan documents will be processed and executed electronically and sent to the sub-registrar’s office, which can also accept the document electronically.

A few charts from the deck showed by Rajeev Chawla are attached at the end. The deck has been shared on “X” / Ex Twitter.

IV. Next Steps

It is imperative that we institutionalise collaboration in digital agriculture – co-creating standards, global pilots & experiments and encouraging cohesive, holistic approach to technology, policy, and funding.

a) Policies to encourage Data Sharing: The need of the hour are critical policy changes to facilitate an “agri-data management framework” being created, to facilitate SOPs on data sharing both by States and the Centre, a mechanism for consent management and grievance redressal. Truly enabling the AgriStack would require several process reforms, and the amendment of procedures, policies and regulations both centrally and across states.

b) Validating technology to protect farmers: The true test of practical impact to a farmer is through ground truthing and testing. It is imperative that new technologies are validated through sandboxes before farmers are exposed to them and minimize the risk that farmers may face when using digital technologies, acting on specific data insights and advisory. At the same time, this should be devoid of heavy Govt / Regulatory oversight and hence the solutions could be co validated, either by the government or by the industry, leveraging Agri universities across agro climatic zones and crop cycles, to at least 70%-80% accuracy..

c) Data curation and accuracy: The panelists emphasised a need to use Open APIs to allow data to be freely shared with consent between the Govt and Private Sector. The framework should be designed to enable data curation, and maintain accuracy, for areas like land records, and real time data. There is also a need to ensure enforcement of AgriStack (like Aadhar was enforced).

V. Topics Uncovered

The panel due to paucity of time, could not cover several, crucial topics about the National AgriStack such as data inclusivity, and the governance of private sector data including addressing security concerns and data breaches.

Further, while monetising data on the data exchange is an incentive for the private sector, we are yet to discuss the several conditions and governance standards under which the private sector would choose to provide their data, consistently, to the data exchange.

This discussion will continue!