Does it make sense to work with a software company in another city?
Many companies in industrial cities such as Bursa, Konya, Kayseri, Gaziantep, Denizli or Manisa struggle to find the right team for a software or AI project locally. Since most of a software project happens on screen, the real question is not where the team sits but how clearly the process is set up and how the result is measured. From Ankara, Hatson Tech runs projects remotely with companies in every city in Türkiye and plans on-site work for steps that need to see the floor.
A remotely run project, step by step
- Discovery: The first sessions are held by video; we map the process, the systems and the success measure together. If needed, we plan an in-depth discovery workshop in person.
- One owner, regular demos: The project has a single technical owner, and we demo a working build every two weeks.
- Test environment: Your team tries every release from its own browser or phone, and feedback is recorded.
- Access and security: Access to your systems is granted with least privilege, logged and time-limited; personal data is processed under KVKK, Türkiye's data protection law.
- Acceptance and go-live: Acceptance tests and go-live can happen remotely or on site; which steps happen where is written into the project plan from the start.
In this model, communication does not scatter across email: decisions, open questions and acceptance criteria live on one project board. From your side, we only need an owner who knows the process and can decide, and a few users to take part in acceptance tests.
AI designed for Turkish data and Turkish regulation
Off-the-shelf AI products from abroad are mostly designed for English. Turkish morphology, in-house language mixed with English terms and scanned Turkish documents can cause unexpected errors in them. We choose solutions by measuring them on a test set built from your Turkish data (Turkish AI and NLP).
On the regulatory side, KVKK and the rules of your sector are part of the design: where personal data is processed, whether it is transferred abroad and how long it is kept are decided up front. If needed, the solution runs entirely on your own servers. We cover the framework in our articles on AI regulation in Türkiye and KVKK and generative AI.
Common needs of companies across Anatolia
| Need | Fitting solution |
|---|---|
| Manual quality control on the production line | Vision-based defect detection |
| Orders and approvals running on spreadsheets and email | Custom software and workflow automation |
| ERP, accounting and field systems that don't talk | Integration and a data warehouse |
| Searching procedures and technical documents | A RAG assistant that cites sources |
| Communication with dealers and field teams | Mobile app |
| Where to start with a first AI project | Short discovery and a measurable pilot |
When choosing a first AI project, keeping the scope narrow and measurable is the healthiest route; we describe our approach in from pilot to production and the cost of an AI project.