The Turkish Artificial Intelligence Language Agents Competition is a national technology competition organized by the Informatics Valley as part of TEKNOFEST 2026, encouraging the development of large language models (LLM), natural language processing (NLP), multimodal artificial intelligence systems, and intelligent agent technologies. The main objective of the competition is to enable participants to develop innovative, applicable, and scalable artificial intelligence solutions focused on the Turkish language for real-world corporate and sectoral problems. In this context, the competition is conducted through three different scenarios: automating public document and correspondence processes with AI-supported intelligent agents; extracting meaningful information from texts related to participation banking products and campaigns using natural language processing methods and making it comparable; and developing multimodal systems that can analyze video data to produce event detection, risk assessment, and decision support outputs. Participants are expected not only to develop working software but also to design AI agents that can perform data analysis, understand context, provide decision support, produce explainable results, utilize open-source technologies, and are applicable in the real world. Thus, the competition aims to contribute to strengthening the Turkish artificial intelligence ecosystem, developing domestic and national technologies, and enabling participants to acquire advanced artificial intelligence skills.
This scenario aims to make document and correspondence processes used in public institutions faster and more efficient through artificial intelligence. The system to be developed is expected to analyze and classify incoming documents, direct them to the relevant departments, and create draft official documents. The primary goal is to develop a document-oriented AI agent with a high level of Turkish language comprehension.
This scenario aims to automatically analyze the campaign and product texts on the websites of participation banks. The artificial intelligence system will extract key data such as profit share rate, maturity, advantages, and campaign information from the texts, organize them, and enable product comparisons. The results are expected to be presented to the user via dashboard and chatbot.
This scenario aims to develop an artificial intelligence system that analyzes videos used in defense, security, and field operations. The system will identify events, individuals, and risky situations in the video; highlight critical moments with timestamps; generate a summary in Turkish; and offer action suggestions to the operator. The solution is expected to operate entirely locally and without reliance on external services.
