An AI-based integrated human-resources analytics and employee security-awareness platform — talent management, candidate scoring and behavioural analytics powered by domain-adapted Turkish NLP/LLM models.
The challenge
HR runs on fragmented data and manual judgement — talent decisions lack objective signals, employee security awareness is hard to measure, and domain-adapted Turkish-language AI tooling for HR barely exists.
Generic language models underperform on Turkish HR text, and off-the-shelf HR suites treat AI as a thin chatbot layer rather than a core capability. For an enterprise that wants measurable, defensible talent decisions, neither option closes the gap.
What we built
STAFİK AI-HRM unifies HR operations around an AI core: a KVKK-compliant data platform, domain-adapted Turkish NLP/LLM models trained on a 2B+ token HR and security-awareness corpus, talent-management modules with a candidate-scoring engine, administrative and security-awareness behavioural analytics, and an intranet portal with productivity and business-intelligence tools.
The programme is structured as five R&D work packages — from data infrastructure and core platform, through NLP/LLM model development and talent modules, to integration, pilot validation and productization — under Türkiye's TÜBİTAK 1501 industrial R&D framework.
The outcome is an HR platform where Turkish-language AI is a first-class capability: candidate scoring with visible reasoning, awareness analytics that quantify risk behaviour, and a roadmap from research to commercial product.