Loading0%
Software Engineer•Industrial AI Builder•Researcher•Product Developer•

Rümeysa Ahsen Çiçek — I build AI systems that work with real operations.

Rümeysa Ahsen Çiçek

Software engineering graduate building AI systems for industry, research, and real products.

(01) About

8
Projects
2
Conference papers
1
Open dataset
Finalist
TEKNOFEST 2025

(02) Selected work

2024 — 2026

Featured case study · Industrial AI

Codlean MES

Turning live factory signals into explainable maintenance decisions.

I built the working AI engine behind an industrial fault prediction system: Kafka streams, predictive models, explainability, an on-premise LLM, five agents and an operator dashboard. The system ran with real customer data before handoff for SCADA screen integration.

98.5%
Model AUC reported in CV
5
Specialized AI agents
In use
Live customer data
Read the case study

Other work

02 — 08 · Scroll to explore

(03) Experience

From model to working system.

The strongest work is measured by what reaches a real workflow. These roles took me from research and model evaluation to software delivered for operational use.

01

Feb — Jun 2026

Düzce, Türkiye

Industrial AI · Client delivery

Atasayar Teknoloji (Codlean)

AI Engine Developer · Co-op

Owned the AI engine of a working MES, from live customer machine data to explainable operator decisions.

98.5%model AUC reported in CV
5specialized AI agents
Livecustomer machine data
  • Built Kafka and FastAPI pipelines, validation and state handling for live machine signals.
  • Developed the XGBoost fault model, SHAP and DLIME explanations, and a local Llama 3.1 assistant.
  • Delivered the working system; SCADA screen integration was the next team step after handoff.
02

Aug — Oct 2025

Elazığ, Türkiye

Computer vision · Edge deployment

Kriptarium LTD. ŞTİ.

Software & AI Engineer · Internship

Compared medical-image models and moved the selected pipeline toward mobile deployment.

20+CNN and ViT models
33DIBaS classes
95.45%accuracy reported in CV
  • Benchmarked CNN and ViT architectures on the 33-class DIBaS dataset.
  • Prepared ONNX Runtime, CoreML and TorchScript paths and built an Android companion app.
  • Joined TÜBİTAK BiGG CUBE with teammates; the application was shortlisted.

(05) Capabilities

The tools follow the problem.

I connect data engineering, models and product delivery so the result works beyond a notebook and makes sense to the person using it.

01 / 03

Industrial data & AI

Live pipelines, risk scoring, predictive maintenance and operator-facing decision support.

  • Python
  • Apache Kafka
  • FastAPI
  • XGBoost
02 / 03

Explainable & deployable models

Evaluating models, explaining their outputs and preparing inference for constrained devices.

  • PyTorch
  • SHAP
  • DLIME
  • ONNX
  • CoreML
03 / 03

Products people use

APIs and interfaces across web, mobile and desktop that bring engineering into a real workflow.

  • React
  • TypeScript
  • SwiftUI
  • Kotlin
  • .NET
Let's Talk★Let's Talk★
© 2026 Rümeysa Ahsen ÇiçekBuilt in Istanbul