Yasemin ARSLAN, MSc.

Founder @ Kairu

About

I completed my Master’s degree in Software Engineering at Celal Bayar University, focusing on Anomaly Detection. Through the company I founded, I provide AI-powered training and consulting services to individuals and organizations, guiding them in developing scalable, sustainable, and business-integrated artificial intelligence solutions. My areas of expertise include artificial intelligence, Python programming, and end-to-end AI system architectures. I have advanced experience in machine learning, deep learning, and large language models (LLMs), covering model design, optimization, performance evaluation, and deployment in production environments. Leveraging the Python ecosystem effectively, I take a systematic approach to data processing, feature engineering, modeling, visualization, and automation workflows. In addition, I work on RAG architectures, vector databases, AI agent development, API design, and MLOps processes. My focus extends beyond model development to ensuring that deployed systems operate reliably, efficiently, and cost-effectively in real-world environments. By balancing performance, accuracy, and sustainability, I help projects generate strategic and measurable value.

Country

Türkiye

City

İzmir

Industry

Computer Software

Skill

Business Marketing, Product Marketing, cı/cd, azure, Continuous Integration and Continuous Delivery (CI/CD), Kubernetes, Jenkins, MLOps, Proje Yönetimi, A/B Testing, Öğretmen Eğitimi, Veri İşleme, Docker, Data scientist, Prompt engineer, Chatgpt, Teaching, Microsoft Office, English, Artificial Intelligence (AI)

Experience

Founder

Kairu

2024-8 - Present · 2 yrs 2 mos

Machine Learning Developer | Artificial intelligence

fybu

2021-10 - 2022-12 · 1 yr 3 mos

İzmir, Turkey

Projects I've worked on: 1- Detection of prices on invoices using algorithmic methods (total, net, VAT amounts). 2- Data preprocessing - Processed data both for price detection and for use in machine learning models. For example, cleaning unnecessary expressions on invoices, standardizing punctuation marks, etc. On the Machine Learning side, I also cleaned and filtered Ngrams in CSV. For instance, I detected that invoice numbers are always on the first page and above half of the page and they must always contain numbers. Based on these findings, I reorganized the data and cleaned the redundancies. 3- Training of machine learning models and hyperparameter analysis - I worked to achieve the best result (f1 score) with many models, compared the results, and selected the best model. 4- Feature Selection - The importance level of features changes for each label. Therefore, depending on which label I was working with, I proceeded by selecting the features that best represent it. This way, I both increased the success and made the system work faster. 5- Web Scraping - Downloading invoices that customers uploaded to the system. 6- Docker 7- Flask - Basic backend-related work.

Electronics and Computer Teacher

Özel Mostem Lisesi

2012-9 - 2021-9 · 9 yrs 1 mo

For 9 years, I worked as an electronics and computer teacher. Throughout this period, in addition to basic electronics knowledge, I conducted lessons with students on programming and transforming projects using development boards like Arduino and Raspberry Pi. I founded the School Drone Team. I facilitated the inclusion of teachers in the certification process for the school's transition to a STEM education approach. Lastly, I spearheaded initiatives to implement Erasmus projects, enabling both teachers and students to gain international experiences.