# A developer from Uzbekistan has created a service for automatic conversion of texts to the new Latin alphabet

Azizullo Temirov, head of the Artificial Intelligence Department at the Ministry of Employment and Poverty Reduction of Uzbekistan, created a free open-source service **[Alifbo](<https://azakapro.github.io/alifbo/?utm_source=thetech>)**, which automatically converts Uzbek texts into the new version of the Latin alphabet. Alifbo is his personal project, unrelated to the activities of the Ministry. The tool works directly in the browser and also allows converting entire Word documents.

The service emerged against the backdrop of Uzbekistan's preparation for the transition to the updated Latin alphabet. It uses new designations for individual Uzbek letters and combinations: **oʻ → ö, gʻ → ğ, sh → ş, ch → ç**.

<img src="https://cdn.the-tech.media/content/2026/09/16/b262ab3e76707899.webp">

The idea for creating Alifbo came to Azizullo after the announcement of the transition to the updated Latin alphabet. In his assessment, this process will require a significant amount of manual work: thousands of pages of documents, websites, educational materials, and archives written in Cyrillic and the old Latin alphabet will need to be adapted.

Existing solutions did not fully meet these requirements: some require uploading files to third-party servers, others disrupt document formatting, are paid, or support only one direction of conversion.

The service supports conversion between Cyrillic, the old Latin alphabet, and the new Latin alphabet in various directions. At the same time, documents do not leave the user's device—all processing occurs directly in the browser. When converting Word documents, the original formatting is preserved: headings, tables, highlights, and footers remain in place. The interface is available in Uzbek, Russian, and English.

The developer paid special attention to cases where automatic conversion might lead to errors. Some letters do not have a one-to-one correspondence between alphabets. For example, the Cyrillic **"ts"** can be rendered as **s** or **ts** depending on the context, and there is no single equivalent for **"shch"**. In such cases, Alifbo highlights ambiguous areas and allows the user to choose the appropriate option manually.

<img src="https://cdn.the-tech.media/content/2026/09/16/76e79e607b8e1863.webp">

**Azizullo Temirov, author of the project** ** Alifbo  
***The alphabet is a common heritage, not someone's product. The tool for working with it should be free and open so that anyone can check the code, fix errors, or integrate it into their own solution.*

**Alifbo can be useful for ordinary users, companies, and government organizations.** Users can convert old documents, diplomas, resumes, family archives, posts, and other texts without the need to retype them manually.

Since the project's code is open, organizations and developers can independently study it, adapt it to their needs, fix errors, or deploy the system within their own infrastructure. Libraries in JavaScript and Python have also been released for developers, allowing them to integrate conversion into their own systems and process large archives automatically.

<img src="https://cdn.the-tech.media/content/2026/09/16/0903f64f9d2e6d5d.webp">

The service works for free and does not require installing additional software.

The service is available [here](<https://azakapro.github.io/alifbo/?utm_source=thetech>).   
The source code is available [here](<https://github.com/azakapro/alifbo>).
