Project Outcomes
Digital access remains a critical barrier for rural women artisans in India’s silk industry. Many face challenges with low literacy, limited familiarity with smartphones, and language mismatches in English-dominant apps. As a result, they often remain dependent on intermediaries, losing both income and autonomy. Resham Sutra, a social enterprise empowering women through sustainable textile livelihoods, partnered with E4C to address this gap by introducing an AI Voice Assistant for Gram Sootra, their mobile marketplace platform.
The project delivered a proof-of-concept voice workflow in English that allows users to list products entirely through speech, replacing “read & type” with “talk & listen.” Packaged as a 40 MB offline module, the system runs on low-end Android devices too and combines Whisper.cpp for speech recognition, fuzzy logic for intent mapping, and Android’s native text-to-speech for guidance. Integrated directly into Gram Sootra, it enables product listing without the need for typing.
The solution represents a significant step toward SDG 5.b – promoting the empowerment of women through ICT. By making Gram Sootra accessible to women with low digital literacy, the assistant enhances economic autonomy, reduces reliance on middlemen, and opens pathways to fairer livelihoods. Looking forward, the system is ready to expand into Hindi and Marathi and extend voice workflows to browsing and purchasing, ensuring more inclusive participation in India’s digital economy.
Key technology/tools used: Java, C++, Python, Fuzzy Logic, Whisper.cpp, Android Native TTS, Figma
Voice Assistant in Gram Sootra. Attribution: Punjaya Wickramasinghe
No Comments.