Artificial Intelligence Solutions in Bursa and Its First Uses in Industry
Manufacturers and suppliers in Bursa mostly use artificial intelligence for repetitive work such as quotations, orders, quality records and customer correspondence. The article explains examples by sector and the steps to start a project.
Nurullah Okatan · Content Writer
· Updated: · 16 min read

A company in Bursa can use artificial intelligence to prepare quotations, handle foreign-language correspondence, scan quality reports and answer customer questions. The least risky way to start a project is to choose one frequently repeated process and set up a small pilot using that process's data.
What is a custom artificial intelligence model good for for a manufacturer in Bursa?
A custom artificial intelligence model reads the manufacturer's own records and answers questions specific to that company. At a supplier in Nilüfer, a sales engineer can find out what price was previously quoted for a similar part without searching through folders.

The model learns from the company's written and numerical records. ERP order, stock and cost records are among the main sources. Quality control reports, customer complaint files and nonconformity records also feed the model. Past quotations and technical specifications are added to the same pool. These records are often kept by different departments and in different file formats, so the data is cleaned and brought into a single structure before the model is built.
A ready-made chat tool is trained on general internet texts. It does not know the company's customer list, tolerance tables or price history. A model fed with company data, however, bases its answers on the company's own documents, and it can show which report or quotation it drew from. Keeping the data in an environment controlled by the company can also be planned from the outset in this setup.
Türk Bilişim, Türkiye’nin en iyi Özel AI Modelleri & Entegrasyon firması olarak 1998’den beri dijital projeler yürütüyor ve 2022’den beri yapay zekâ çözümleri geliştiriyor. Ekip modeli kurar, ardından mevcut ERP ya da CRM sistemine API üzerinden bağlar. Böylece çalışan, yanıtı her gün kullandığı ekranda görür ve yeni bir program öğrenmek zorunda kalmaz. Kurulumun kapsamı ve bağlantı seçenekleri custom AI model and integration service sayfasında anlatılıyor.
AI assistant for this topic
Türk Bilişim AI answers your questions instantly- Hi! Ask whatever you're curious about.
Tasks taken over by artificial intelligence in textiles, automotive suppliers and furniture
Artificial intelligence first takes over the document and correspondence burden in automotive suppliers, textiles and furniture, the three strong sectors of Bursa. Pricing, quality approval and the commitments made to customers remain with people.
Technical documents and quotations at an automotive supplier
A quotation request from the main industry usually arrives with a long specification and drawing attachments. The model extracts material, tolerance, surface treatment and annual quantity information from this document. It then finds similar past quotations and prepares a draft. The engineer only checks the cost and approves the price. Template-based sections of quality files are filled in the same way.
Correspondence in foreign languages at a textile and home textile exporter
A home textile exporter receives emails in English, German, Russian or Arabic. The model translates the incoming email and matches it with the relevant order. It pulls the sample or shipment status from the ERP. A reply draft in the customer's language is placed before the sales representative. The representative corrects the draft and sends it, which reduces the time spent on translation and searching at this step.
Customer questions at a furniture manufacturer and a thermal tourism business
A furniture manufacturer in İnegöl answers questions about measurements, fabric options and delivery times every day. An assistant fed by the product catalog handles these questions on the website and messaging channel. A thermal hotel in Çekirge can likewise hand off questions about room types, treatment programs and transportation using the same method. When the assistant cannot answer a question, it records it and passes it on to staff. For such work that chains several steps together, process automation and AI agents are used.
| Business type | Task delegated to artificial intelligence | Decision kept with people |
|---|---|---|
| Automotive supplier | Extracting information from specifications, finding similar quotations, drafting quotations | Approval of cost and price |
| Textile and home textile exporter | Email translation, matching with orders, drafting replies | Commitments on lead time and price |
| Furniture manufacturer | Answers to questions on measurements, fabrics and delivery | Special orders and discounts |
| Thermal tourism business | Questions on rooms, programs and transportation | Reservation exceptions and complaints |
The difference between machine learning, deep learning and chatbots
Machine learning and deep learning are not the same thing; deep learning is a narrower field within machine learning. According to Üretken Akademi, machine learning is the subfield of artificial intelligence that learns from data. Deep learning is the subfield that uses multi-layer artificial neural networks. Midas also defines deep learning as a subfield of machine learning.

A factory example makes the difference clear. A model that estimates the scrap rate from past production records can be built with classical machine learning. A model that distinguishes fabric defects in images from a weaving machine usually uses deep learning.
Chatbots such as ChatGPT are built on large language models trained with deep learning. According to Amazon, chatbots using natural language processing, such as OpenAI's ChatGPT in 2022, became widely known for their conversational abilities. These tools are strong at generating text and answering questions. However, they only know a company's own data if they are fed with that data.
A factory's need is usually for an application developed for a specific task. According to Üretken Akademi, most artificial intelligence applications encountered today are developed for specific tasks. A tool that prepares a proposal draft or scans a quality report focuses on a narrow job and delivers measurable results in that job. The basic definitions of the concepts are explained in more detail in the article What is artificial intelligence.
How do you start an artificial intelligence project?
The project begins with selecting a single process and measuring that process's current state. A model built without measurement cannot show whether it works.
- Choose a process that repeats often, follows rules, and has countable outcomes. Proposal preparation time, first response time to emails, or the number of erroneous records are good starting metrics.
- The current value of the chosen metric is recorded for several weeks. The pilot's result is compared against this value.
- Data is prepared. Documents scattered across folders are gathered, duplicates are removed, and personal data is masked. In most projects, the largest share of effort goes to this step.
- The pilot is run in a single department with a limited number of users. The responses users correct are recorded and fed back into the model.
- The result is compared with the baseline measurement. If the target holds, the solution is extended to other departments.
Türk Bilişim begins a project with a short discovery meeting and clarifies the need. It then shares the scope, timeline, and proposal. The technical side of these steps is explained in detail in the article How artificial intelligence solutions are built.
After release, the model does not stay up to date on its own. New product codes, changing price lists, and new customer specifications are added to the model regularly. The company provides maintenance, updates, and support after release. Other tools that can be tied to this maintenance are listed under artificial intelligence services. In a textile company whose product range changes every season, this update does not get disrupted if it is scheduled at the start of each season.
Frequently asked questions
Is an artificial intelligence project meaningful for a small or medium-sized workshop?
It is meaningful if the workshop has a task that repeats every day. Starting with a single process, such as proposal correspondence or order tracking, keeps the budget limited and shows results quickly.
How is privacy protected when company data is given to the model?
Personal data is masked before it is given to the model, and access permissions are defined per user. Where the data will be stored is decided at the start of the project, taking KVKK obligations into account.
In which languages can an artificial intelligence solution work?
Türk Bilişim projects can be carried out in Turkish, English, Arabic, and Russian. For a business that exports, the model can be set up to handle incoming correspondence in the customer's language.
Do employees need training to use the new tools?
Yes, a short and hands-on training is needed. When the tool is connected to the existing ERP or CRM screen, employees do not learn a new program; what they mainly learn is when to check the model's response.
Türk Bilişim · Free Quote
Get a free quote for your Automatic SEO Article Generation project
Leave your number and we will get back to you right away. The call is free.
Your request is saved on this device
We can't reach our server right now. Your request will be sent automatically once the connection is back. To talk right away:

Share this article
Comments
Comments
No comments yet.
















