Purpose: This paper aims to elaborate the library data migration process from LAMP (Library Automation Management Program) to the open source software Koha's (2.2.8 Windows based) Pakistani flavour PakLAG-Koha in six legislative assembly libraries of Pakistan. Design/methodology/approach: The paper explains different steps of the data migration from one ILM (integrated library management system) to a new ILM in six libraries over the course of three months. Qualitative research methods, i.e. direct observation and interviews, were adopted for finding the problems and solutions of the delayed project. Training and support was offered by a local volunteer group (PakLAG). One of the authors of the paper presented first-hand experience of being a consultant of this project. The study and results are the author's own reflections and first-hand experiences, and hence results are based on direct observations. Findings: The paper identifies several critical issues and their solutions concerning data migration within a local scenario as well as how local consultants and support groups are coping with these issues. It was noticed that the data migration can be done quickly if any support group is properly involved with the library staff. Originality/value: This paper provides first-hand experience and success story of library data migration from DOS-based ILM to open source software. Process, tools and findings discussed in this paper are valuable for the libraries planning to convert non-MARC data to open source MARC-based software Koha. (Contains 4 figures and 1 table.)