In this paper, an extractive text summarization system is proposed to be formatted for Azerbaijani language using modern natural language processing (NLP) approaches. Although there are tremendous advancements in travelling from English to other languages like French and Spanish, unfortunately the same could not be said about Azerbaijani; the main reasons are that Azerbaijani has a rather rich morphology and lacks proper resources. BERT embeddings have been utilized to represent sentences, while the TextRank algorithm has been employed to rank them based on their relevance. Raw text is transformed in the system through tokenization, then BERT based sentence embedding is applied, similarity matrix is constructed and TextRank algorithm is applied for sentence ranking for the purpose of summary generation. Furthermore, there is also a possibility of rule-based filtering which is used to avoid repetition and to create the final summaries more systematically. The evaluation results, measured in terms of precision, recall, and F1 score, were found to be competitive when benchmarked using the ROUGE metric. This work can be viewed as a starting point for further research in Azerbaijani NLP and is made with the intent to apply improvements to model in the future, in terms of enhancing precision in the generated summary using abstractive summarization.