Credit bureau and AI infrastructure platform in Turkish financial services
KKB is Turkey's primary credit information bureau, serving 180+ financial institutions, individuals, and businesses since 1995. The tech stack reveals a sharp pivot toward AI infrastructure: heavy Python/PyTorch/LangChain tooling with vector databases (Qdrant, Weaviate, Pinecone, Faiss), RAG pipelines, and agentic AI applications dominate active projects. This contrasts sharply with legacy data-center infrastructure work (SCADA, cabling), signaling an organization in mid-modernization—building AI capabilities while maintaining mission-critical credit-reporting systems.
KKB (Kredi Kayıt Bürosu) operates as Turkey's Risk Center, the sole information service provider for financial data sharing across Turkish banking and finance. Founded in 1995 with 9 bank partners, the bureau has grown to serve over 180 member institutions including banks, consumer finance companies, leasing firms, factoring operations, and insurers. In 2013, KKB expanded beyond institutional clients by launching an electronic reporting system, opening services to individuals and non-financial businesses. The organization operates in the 501–1,000 employee range and is based in İstanbul.
KKB is actively building with Python, PyTorch, and Transformers. Active projects include RAG pipelines, vector database schema design, and agentic AI applications using frameworks like LangGraph and CrewAI. Vector databases in use include Qdrant, Weaviate, Pinecone, Faiss, and LanceDB.
KKB faces vector database design, model deployment/MLOps, dataset challenges, and building security infrastructure from scratch. The organization is also updating existing systems while scaling agentic AI applications and end-to-end AI services.
Other companies in the same industry, closest in size