Project TitleThe Use of NLP and DLT to Enable the Digitalization of Telecom Roaming Agreements





Why is Telecom Roaming Agreements?

Telecom roaming agreements are legal and business documents constructed based on industry standard templates provided by the GSMA organization. The roaming feature in mobile networks enables a mobile subscriber to automatically utilize the visiting network resources to either make and receive voice calls, send and receive data, or to access other services when travelling outside the geographical coverage area of their home network. For communication service providers to be able to exchange those services and be able to establish the business relationship and maintain the billing records, a legal dedicated roaming agreement need to be drafted, negotiated and executed.  This agreement will contain the legal and commercial terms, which is usually bilateral. For a communication service provider the process o drafting and negotiation for those bilateral agreements will be time consuming and error prone and facilitating the process of great benefit. The GSMA association has started the effort to make this process not repetitive and build on the previous knowledge by providing a set of template. However, a digitized copy of those drafting articles with a configurable set of variables and their different variants can enable to digitalize the entire process of drafting and negotiation to help in speeding up the process as well as minimizing the errors and risks for CSPs.

What is this project about?

The project looks at how to facilitate the process of Telecom roaming agreements drafting and negotiation. The project looks at first constructing a library of drafting articles with a set of variables that will be extracted from the available templates and previous roaming agreements using Natural Language Processing process NLP. The second part will be towards translating the drafting and negotiation process as a chaincode on Blockchain to digitalize the process and provide a maintainable and actionable copy of the agreement. The project focus on how to be able to digitalize the legal process and make it automated through smart contracts to be more efficient and less error prone.

What is the overall project suggested plan?

The suggested project plan will be conducted in three phases. The first is to collect and gather the main GSMA templates and previous agreements from CSPs, analyze those agreements text and define the main features to build the NLP model. The second phase will be focused on mapping the drafting and negotiation process into the smart contract logic that will enable at the end constructing the draft agreement utilizing the NLP model constructed from the first phase. The third phase will be towards building a simple UI that will handle demonstrating building a simple agreement utilizing the outcome of the first and second phase.

Additional Information

Learning Objectives

1. Provide understanding of Hyperledger fabric concepts

2. Provide a learning experience on how different open source and industry standardization efforts can align. 
3. Learning to manage (coding/installing/executing) Hyperledger Fabric chaincode.
4. Provide good exposure to the use of DLR technology in real life use case
5. Enhance the knowledge on NLP and its use in legal documents digitization.
6. Gain Hands-on experience with opensource software development

7. Provide a unique experience on how different technologies and domains can integrate to solve real life problems

Expected Outcome

The expected outcome of this project, would include:

  1. Building a drafting library and an NPL model that will include a digital version of the most relevant and important articles in a the GSMA AA.12, AA.13 and AA-14 roaming agreement templates.
  2. Developing a chaincode that will map all steps of the drafting and negotiation business processes for roaming agreements.
  3. Building a PoC for UI to handle the agreement construction
  4. Writing a solution document summarizing the implementation details

Relation to Hyperledger 

The expected outcome provides a very appealing use case for the how Hyperledger Fabric can enable automating the business process of roaming agreements in Telecom. Apart from the drafting and negotiation process the proposed approach can help in shoing how this can be extended to work in other verticals for example in supply chain and possibility later on for integration with Baseline protocol.

Education Level

Masters or Ph.D. level students are preferred.



  • Willingness to contribute to a meaningful mission, in an open-source mentality
  • Teamwork skills, and skill to maintain good communication with other parties that might be involved to ensure a successfully completed project
  • Good understanding of Hyperledger Fabric technology
  • Experience with one the following programming language for chaincode development (javascript or go ) 
  • Experience with building NLP models.

Nice to have:

  • Telecom related experience.

Future plans

The plan for this project is to be incubated in the GSMA or TM Forum catalyst projects to make it go into a more formal industry standardization, and also to extend that to Service Level Agreement SLA automation.

Preferred Hours and Length of Internship

Both Full-time or Part-time are possible options, preferably full-time.

Mentor(s) Names and Contact Info

Ahmad Sghaier Omar,

Mohamed Elshrif,

Noureddin Sadawi,


Santiago Figueroa-Lorenzo CEIT and University of Navarra,,

Project Results

Medium Articles that support the project

The project has been documented through the following Medium articles:

  1. Blockchain-based digitization of the roaming agreement drafting process
  2. NLP Engine to detect variables, standard clauses, variations, and customized texts
  3. Chaincode design for managing the drafting of roaming agreements
  4. Chaincode implementation for managing the drafting of roaming agreements

Scientific Contributions that support the project

  1. A Natural Language Processing Approach for the Digitalization of Roaming Agreements
    • Sent to the Conference: ILCICT 2021 (Current status): Under review

Relevant Project repository:

Final Report


NLP Part:

NLP Part.mp4

HFB Part:

HFB Part.mp4