From idea to intelligent product.

Who we are

Hatson Tech is an AI and software development company founded in Ankara in August 2026. We build AI agents, RAG assistants that cite their sources, fine-tuned LLMs, computer vision, data engineering, and web and mobile applications for businesses.

Most AI projects succeed or fail in the software around the model rather than in the model itself: data preparation, permissions, integrations, measurement and maintenance. That is why we brought data analysis, AI engineering and full-stack development into one team. The people who design the model also deliver the product that uses it; we do not hand the work over to another team.

We work with clients in Türkiye and abroad, in English and Turkish. Projects are run from Ankara, and when needed we deploy on the client's own servers, on premises.

What we do

Each service has its own page describing our method, how we measure results and common questions:

• AI agent development: agents connected to business systems, with human approval and audit trails.
• RAG and enterprise semantic search: assistants that answer from company documents and cite them.
• LLM training and fine-tuning: Turkish and domain-specific language models.
• Computer vision: quality inspection, OCR and document analysis.
• Data engineering: pipelines, warehousing and data quality.
• Custom software development: web applications, dashboards, APIs and cloud infrastructure.
• Mobile app development: iOS and Android apps.

For companies whose data cannot leave their premises, we build on-premise enterprise AI.

How we work

1. We validate the problem first. Not every task needs AI. If simple automation or an existing tool will do, we say so.
2. We make success measurable from the start. Before a pilot begins we agree on the success criteria and target values in writing, and we measure results on evaluation sets built from real cases, not on impressions.
3. Privacy is part of the design. Which data goes where, who can see what, and KVKK/GDPR obligations are settled in the first week of the project.
4. We start small and grow step by step. We begin with a narrow pilot and widen the scope as it proves its value.
5. We stay after delivery. We hand over the code, documentation and training, and if you want, we keep maintaining and extending the system.

Our writing

In our Insights section we publish practical guides on AI agents, RAG, data and software architecture. Every article names its author, cited claims link to primary sources, and updates are dated openly.

Company details