Funding

India's food labelling system carries one of the most linguistically diverse ingredient landscapes in the world — the same substance appears under dozens of names across regulatory filings, procurement systems, and consumer labels, across languages and scripts. This isn't a data quality problem. It's the natural result of a food culture that is genuinely plural. The problem is that existing digital systems treat this diversity as noise to be standardised away.

The Indian Food Informatics Data (IFID) project is building a coordination layer that doesn't require that trade-off. The goal is a shared infrastructure where thousands of regional ingredient expressions — Mathania red chilli, Kashmiri mirch, lal mirchi powder — all map to stable, machine-readable identifiers without any one of them being declared the "correct" name. Brands keep their regional identity. Regulators get a consistent reference. Researchers get reproducible data. The infrastructure works for all of them because it coordinates without converging.

This matters beyond food. It is a demonstration that the design assumption embedded in most data standardisation projects — that coordination requires convergence — is not inevitable. It's an infrastructure choice. IFID is building the alternative.

Everything we produce is released openly under CC BY 4.0 — datasets, schemas, classification models, reports. The work is public infrastructure, not proprietary tooling.

Published under the IFID project so far, all open access on Zenodo and Kaggle:

• Dataset
  Indian Food Ingredients & Label Variants 2,500+ regional label variants mapped to canonical identifiers. The ground-truth dataset for calibrating extraction models against the actual diversity of Indian ingredient naming.
• Report
  Regulatory Delta of Food Labelling Laws in India: A Comparative Analysis of the FSSAI 2011 and 2020 Regulations This report summarises the regulatory delta between the Food Safety and Standards (Packaging and Labelling) Regulations, 2011 and the Food Safety and Standards (Labelling and Display) Regulations, 2020.
• Report
  Identity, Transformation, and Function: A Tri-Axial Model for the Classification of Food Ingredient Identity A principled framework for assigning a determinate identity to any food ingredient across three axes — processing energy, matter state, and functional role.
• Report
  Indian Supreme Court Defines Hierarchical Classification Framework for Food Products, Overruling Common Parlance Precedents This report synthesizes landmark judicial decisions in Indian food classification law, documenting the transition from common parlance-based interpretation to a hierarchical technical classification framework.
• Paper
  Encyclopedia of Indian Food Ingredients 600+ food components structured in JSON, Markdown, and LaTeX. Machine-readable, multilingual, interoperable with Open Food Facts and FSSAI.
• Data Paper
  Justification Companion to the EMF-Scoring Model The evidentiary backbone for the EMF model — chemical justification, regulatory mapping, and ITC-HS classification for each scoring decision. Designed to be audited and challenged.

Further validation — stress-testing the classification logic against edge cases, engaging domain experts across food science, regulatory law, and trade — is the immediate next phase. That is where support would have the most direct impact.

We are an intentionally small team working fully remotely. There is no office to maintain. Support goes to four things and nothing else:

• Infrastructure Servers, storage, and hosting for datasets, APIs, and research pipelines. The IFID infrastructure needs to stay available and reproducible — these are costs that recur regardless of the pace of publication. Ongoing
• Systems API access and compute for large-scale constrained extraction, LLM-assisted classification, and audit pipelines. This is the tooling cost of doing the work at the scale India's food system actually requires. Ongoing
• Researcher Stipends Stipends for researchers working on active projects. We keep the team small deliberately — two researchers on IFID currently — and we pay fairly for that commitment. Per project
• Domain Expert Fees Fees for subject matter experts brought in to validate and stress-test the logic — food scientists, regulatory specialists, legal experts. Getting the classification right matters more than publishing quickly, and that requires people with real domain depth. Per engagement

Acknowledgment is offered in proportion to support. Publication naming is reserved for contributions that meaningfully shape the lab's capacity to operate.

• I Friend
  Friend of the Lab Up to ₹25,000

  • Named in the acknowledgments of the relevant project's Zenodo record
  • Named in the lab's annual acknowledgments page on the website
• II Supporter
  Project Supporter ₹25,000 – ₹1,00,000

  • Everything in Tier I
  • Named in the acknowledgments section of all publications produced during the supported period
  • Updated by email when new outputs are published, with a brief note on what they cover
• III Institutional
  Institutional Partner ₹1,00,000+

  • Everything in Tier II
  • Organisation named as a supporting institution on the iSRL website and in all publications during the supported year
  • Logo on relevant project pages and published datasets
  • Informed first when the next year's project focus is scoped

Research groups and labs with domain expertise in food science, regulatory systems, nutrition, or related areas are welcome to collaborate directly — contributing to validation, red-teaming the classification logic, or identifying where the framework breaks.

This kind of collaboration is distinct from financial support. It's a contribution of expertise, and it is acknowledged as such in all relevant publications. If your lab works in a domain that intersects with this problem, we'd genuinely like to hear from you.

There is an option to sponsor a specific researcher's stipend directly — covering the cost for a researcher to work on an active iSRL project for a defined period.