Leveraging LLMs and supporting entrepreneurship with Dialog

Imagine being able to create an app by simply chatting to “someone” in your native language…that by asking questions and giving commands, you can create a simple app without having to know or write a single line of code.  

Well as part of a ‘proof of concept’, we have shown that Dialog Sri Lanka’s Ideamart AppMaker, an initiative supported by the GSMA AI4I and newly launched M4D Emerging Tech programme, will allow users to do just this: to create simple Android apps using ready-made templates— including uploading images and content – without needing to code. 

The context  

In Sri Lanka, although women have a high literacy rate, only about 30% participate in the workforce. As of 2017/18, just 25% of SMEs – the backbone of the economy – were women-owned, with barriers including cultural norms, limited financial and digital access, and a lack of tech support. Identifying this gap, and working to address it, Dialog launched IdeaMart for Women which provided low-costs tools, mentorship and technical support for women entrepreneurs.  

From this, and as technological solutions accelerated and AI became ubiquitous, Dialog spun out Ideamart AppMaker, to attempt to continue to lower barriers to entry – especially for women – and to help close the digital divide as AI becomes more prevalent. 

Current AI Limitations 

Currently, an important issue with AI and specifically large language models (LLMs) is their limited availability of local languages which can risk widening or exacerbating the digital divide. Currently in Sri Lanka, some 16 million people (over 70% of the total population) speak Sinhala as their first language and 5 million speak Tamil, meaning that a local-language LLM is critical for accessibility. 

The Solution 

Identifying this as a key barrier, the GSMA worked with Dialog over three months to develop a proof of concept that integrates large language model (LLM) into the Ideamart AppMaker platform. This included work on leveraging large language model (LLM) in a local language that enables users to describe their app idea in everyday language and receive a functional app in return. The technology enables users to describe the app features and functionalities that they require in their native language. These are then “translated” into technical, platform specific commands that can: allow app template selection, modify the appearance of the selected templates, execute essential features of the template and integrate images and content. Essentially, this allows the platform to comprehend written commands in local languages with the output being a functional app that required no technical skills to develop.   

The Outcome 

The proof of concept has been incredibly successful with positive feedback from Dialog and end users alike. One end-user said; “The AI tool itself has a lot of potential, and it’s really impressive how it simplifies the process of app building. The linguistic capabilities of understanding Singlish and ability to converse in the same tone is amazing…” . Another said: “The platform is a great step forward for non-tech-savvy users who dream of starting online businesses… I was impressed by the AI’s ability to understand all three major Sri Lankan languages, even when mixed”.  

Comments like these demonstrate just how valuable these new AI tools can be and how critical they will be to helping reduce barriers to entry, especially for women and particularly in areas that have previously required more technical knowledge and understanding. With these new AI tools, there are growing examples of how AI can be used to close the digital divide further.  

Dialog is now finalising UI iterations and testing, with plans to move into production later this year. And with a further aim to replicate this proof of concept in additional countries utilising the lessons learned.  

If you would like to learn more about this project, or to work with us, please get in touch at betterfutures@gsma.com


This initiative is currently funded by UK International Development from the UK government and is supported by the GSMA and its members.
The views expressed do not necessarily reflect the UK government’s official policies.

Logo with a stylized UK flag on the left. To the right, blue text reads "UK International Development," highlighting our commitment to leveraging LLM technology. Below, smaller text says "Partnership | Progress | Prosperity." Clean and modern design.