AI for All Challenge

AI for All – India’s Open Data & AI-Readiness Challenge AI for All is a nationwide online hackathon by Factly in collaboration with Meta, aimed at identifying innovative, open-source solutions that make India’s datasets and digital resources ready for responsible AI use—bridging data, tools, and inclusion. The hackathon invites students, AI startups, researchers, and civic-tech innovators to build solutions across data readiness, Indic language enablement, public data extraction, and AI for social impact. Shortlisted teams will receive mentorship and present their work to an expert jury. All final solutions will be released as open-source and linked to AIKosh, ensuring long-term public value and reuse.

Detailed Themes & Focus Areas

1. AI for Social Impact

Low-code or no-code workflows leveraging open data to build solutions for health, environment, education, or MSME growth, demonstrating AI’s societal value.

Example Projects

AI for Social Impact

Title: Multilingual Scheme Chatbot for myScheme Users

Problem: Citizens often don’t know which government schemes they qualify for due to language barriers and complex portals.

Project: Build a WhatsApp/Telegram chatbot that uses myScheme data/APIs to answer eligibility questions via voice/text in local languages.

Title: AI Toolkit for Anganwadi Performance Dashboards

Problem: ICDS Anganwadi data (nutrition, attendance, growth charts) is fragmented.

Project: Build an AI-powered dashboard that flags at-risk children, predicts dropout risk, and summarises village-level trends.

2. Data Readiness & Standardisation:

Tools for automated cleaning, metadata extraction, schema harmonisation, and interoperability of public datasets to make them AI-ready and reusable across domains.

Example projects

Data Readiness & Standardisation

Title: Unified Metadata Extractor for Indian Government Portals

Problem: India’s government portals & data portals may use inconsistent metadata and poor documentation.

Project: Build an automated metadata extractor + harmoniser that standardises dataset

titles, descriptions, formats, temporal ranges, and geotags into a unified schema compatible

with AIKosh.

Title: Census & Survey Interoperability Engine

Problem: Census, NSSO Reports, NFHS, and other survey datasets differ greatly across years in schema, terminology, district boundaries, and formats.

Project: Build an engine that aligns multi-year Census & NSSO indicators, normalises

geographic units, and outputs time-series AI-ready datasets with consistent variable names and units.

3. Indic Language Enablement

Models and tools for multilingual text, speech, and image extraction, transliteration, synthetic data creation and translation that expand access to AI in India’s diverse linguistic ecosystem

Example Project

Indic Language Enablement

Title: Regional Language OCR & Corpus Builder (e.g., Regional language Newspaper archives)

Problem: Huge libraries of scanned vernacular newspapers and books remain un-digitised,

limiting NLP development.

Project: Build a high-accuracy OCR pipeline for Indic scripts to convert scanned archives

into clean text corpora for model training.

4. Public Data Extraction & Structuring

Automated pipelines for extracting, cleaning, and classifying information (tables, charts, text) from PDFs, reports, and archives to make government and institutional data AI-ready

Example Projects

Public Data Extraction & Structuring

Title: Government Tender & Procurement Normalizer (GeM + CPPP + Railways)

Problem: Tender datasets from GeM, CPPP, Railways, and state procurement portals use different formats, terminologies, and document styles, preventing unified procurement analytics.

Project: Build a pipeline that extracts tender details from PDFs/webpages (items, quantity,

price, department, vendor details) and standardises them into a single schema with harmonised fields and deduplicated entries across platforms.

Title: Infrastructure Project Tracker from PAIMANA website

Problem: Progress reports of major infrastructure projects (roads, railways, irrigation) are available on PAIMANA portals but are not intuitive for user information

Project: Create an AI system that extracts details like milestones, geo-locations, and progress percentages from monthly progress reports, producing a clean, structured database & interface that allows tracking cost overruns and delays.