Background of the Problem Statement
A large share of India's workforce has acquired trade skills informally, through apprenticeship-style on-the-job experience rather than structured, certified training and has no formal credential to show for it. NCVET's Recognition of Prior Learning (RPL) framework exists precisely to certify such workers against NSQF levels without requiring them to repeat training they have effectively already completed. In practice, RPL assessment still depends heavily on manual practical evaluation by assessors, which is slow to scale, inconsistent across assessors and locations, and difficult to schedule for workers who cannot easily take time off. A structured, AI-assisted assessment tool that can evaluate an informal worker's practical competence against NSQF-aligned criteria combining structured self-declaration,practical task evaluation aids, and assessor-support scoring, would let RPL assessment scale reach far more of the informal workforce than manual-only assessment currently allows.
Description of the Problem Statement
The challenge is to build an AI-assisted RPL assessment tool that can:
Guide a worker through a structured self-declaration of prior experience, mapped automatically to the closest relevant NSQF qualification pack(s). Support practical skill evaluation through guided task checklists, and where feasible,video- or image-based assessment aids that help an assessor score consistently against NSQF criteria. Standardise scoring rubrics across assessors and locations to reduce evaluator-to-evaluator variance in RPL outcomes. Generate an NSQF-aligned competency profile and certification recommendation for assessor sign-off (the tool supports, but does not replace, the human assessor's final decision). Work in low-connectivity settings, with offline data capture and later sync, given that much of the informal workforce is in semi-urban and rural locations.
Expected Solutions / Outcomes
A working assessment workflow covering self-declaration through an assessor-facing scoring interface for at least one trade. An NSQF qualification-pack mapping engine. Evidence of consistency improvement over unassisted manual scoring (e.g., inter-assessor agreement on a test set). An offline-capable mobile/web interface. A clear description of where the tool supports versus replaces assessor judgement, to preserve certification integrity.