I'm a computer and AI engineer, a graduate of METU Computer Engineering. I build genuinely useful products with machine learning and natural language processing (NLP) — these days at EcoFluxion.
Computer and AI engineer
I'm a computer and AI engineer, a METU Computer Engineering graduate. In short, my work is turning machine learning and natural language processing (NLP) into products people can rely on.
I care about taking an idea and turning it into an end-to-end, scalable product on a solid engineering foundation.
I work with machine learning, natural language processing (NLP) and language models — carrying a model from experiment to production.
I care that models don't just work, but are accurate, measurable and trustworthy.
In short: I take a machine-learning or AI idea and turn it into a product people can trust and use. Most of my work goes through the same three steps.
First I adapt the data and the model to the problem at hand. Then I ground the answers in real sources with RAG to make them reliable. Finally I move the model into production and keep it observable and scalable.
These days I do this at EcoFluxion — working on İçtiHub, a legal product whose idea I helped pioneer.
The engineering behind the products. All of it serves a single purpose: accurate, fast, and reliable Turkish artificial intelligence.
She trains and deploys her own models fine-tuned to the Turkish legal language.
Understanding Turkish and multilingual text: classification, inference, and meaning-based processing.
A retrieval architecture that prevents the model from “making things up” and grounds answers in real documents.
Fast, meaning-based search rather than keyword matching. Accurate retrieval across large archives.
Agent architectures that perform smart routing and use tools via the Model Context Protocol.
Training, fine-tuning, and continuous deployment on large-scale infrastructure. TRUBA and HPC experience.
In-depth, research-based writing on artificial intelligence, legal technology, and startups.
An intuitive guide to CNNs through filters, convolution, pooling and layers: how a machine actually parses and "sees" an image, with a short code sketch.
Read
An intuitive guide to the evolution of sequence architectures: how RNNs work, their limits like vanishing gradients and the inability to parallelize, the gates of the LSTM, and why attention won out with the Transformer.
Read
Why do models memorize? An intuitive guide to overfitting and regularization: dropout, L1/L2, early stopping, and cross-validation explained with everyday examples.
Read
An intuitive walk through gradient descent via the loss function, learning rate, SGD and Adam: how does a model really "learn" by measuring and reducing its error?
Read
Starting from the confusion matrix, we explain accuracy, precision, recall, F1, and ROC-AUC through everyday examples and clarify which metric is the right choice and when.
Read
An intuitive guide to unsupervised learning through k-means, hierarchical clustering and PCA: how to extract hidden patterns from unlabeled data and choose the number of clusters.
ReadThe turning points of a story that runs from engineering to building products.
She gained a solid engineering foundation in Computer Engineering at Middle East Technical University, learning the depths of algorithms, computer science, and system design.
She accumulated deep, years-long engineering experience in computing and artificial intelligence, mastering system architecture and product development.
Starting from a real need, she became one of the pioneers of the İçtiHub idea; she turned a problem into a product idea that would add value for legal professionals.
She brought EcoFluxion Teknoloji A.Ş. to life as a co-founder at ODTÜ Teknokent in Ankara. The goal: its own Turkish-focused artificial intelligence products.
She is turning İçtiHub into a complete product with an LLM and RAG architecture purpose-built for the Turkish legal language, growing the team together with co-founder İsmail Tarık Şenkal.
If you'd like to talk about artificial intelligence, products or a possible collaboration, don't hesitate to write.