Skip to content
AI / ML EngineerComputer EngineerEnd-to-End Developer

I turn complex problemsinto elegantsystems.

From model training to production deployment, I design end-to-end systems. In the customs & logistics sector, I build AI-powered automation infrastructure, enterprise software solutions, and data-driven products — at every layer, from the backend to mobile.

Emre Canik

Emre Canik

Istanbul, Turkey

Open to work

To solve a problem, I pick the right technology and deliver it as an end-to-end product. From the OCR engine to the declaration portal, from the RAG pipeline to the mobile app — I write every layer that carries an idea into production.

I can train models in Python and write enterprise integrations in C#. I can build dashboards with React and ship them with Docker. I'm not tied to a single stack — I use whatever tool best fits the problem.

“Good engineering starts with asking the right question — and ends with choosing the right technology.”

0Years of Experience
0Documents Processed
0OCR Accuracy
0Live Products

Technologies.

Machine LearningDeep LearningComputer VisionNLPEnd-to-End DevelopmentC# / .NETDevOpsMLOpsLLMOpsReactNext.jsPythonTypeScriptPyTorchDockerKubernetesRustOCRMachine LearningDeep LearningComputer VisionNLPEnd-to-End DevelopmentC# / .NETDevOpsMLOpsLLMOpsReactNext.jsPythonTypeScriptPyTorchDockerKubernetesRustOCR
OCRRustKubernetesDockerPyTorchTypeScriptPythonNext.jsReactLLMOpsMLOpsDevOpsC# / .NETEnd-to-End DevelopmentNLPComputer VisionDeep LearningMachine LearningOCRRustKubernetesDockerPyTorchTypeScriptPythonNext.jsReactLLMOpsMLOpsDevOpsC# / .NETEnd-to-End DevelopmentNLPComputer VisionDeep LearningMachine Learning

AI / ML

PyTorchTensorFlowScikit-learnXGBoostPandasOpenCVHugging FaceLangChainONNXComputer VisionNLPRAG

Frontend

ReactNext.jsTypeScriptTailwind CSSFramer MotionHTML5 / CSS3

Backend

PythonC# / ASP.NETNode.jsFastAPIRustMS SQLPostgreSQLMongoDBRedisREST APIGraphQL

DevOps / MLOps

DockerKubernetesAWSCI/CDGitLinuxTerraformMLflowNginx

Mobile

React NativeExpoFlutterFirebase

Experience.

AI automation, enterprise software, and international freelance projects in the customs & logistics domain.

2024
Jun 2024 — Present

AI / ML Engineer & Software Developer

Kızılkaya Gümrük Müşavirliği A.Ş.

Istanbul, Turkey Full-time

94%

OCR Accuracy

50K+

Documents Processed

6+

Products Live

I'm building an end-to-end, AI-powered automation ecosystem for the customs & logistics sector. I develop full-stack solutions ranging from systems that cut a declaration process taking an agent 2 days down to 1 minute, to control mechanisms that prevent million-dollar penalties.

DOEP: an AI-powered platform for drafting and validating customs declarations and querying customs systems — reduced a 2-day manual process to 1 minute, automatically detecting and preventing million-dollar customs penalties
AI Project & Task Management Portal: automatic task distribution, assignment suggestions, workload analysis, and workflow optimization
Rosetta 3.0: extracting structured JSON from multi-page PDFs/documents with OCR + LLM — 94% accuracy
Replicant: semantic search over 10,000+ pages of regulations with a RAG architecture — response time 12s → 0.8s
Kızılkaya Mobile App: customs operations tracking and notification system
SQLVision: an enterprise data analysis and reporting tool built on MS SQL + stored procedures
Enterprise automation with C# / ASP.NET MVC, plus e-Invoice/e-Declaration integrations
Production deployment with Docker/K8s on an EPYC server, plus 4x inference acceleration without a GPU via ONNX INT8 quantization
Python
C#/.NET
PyTorch
Transformers
ONNX
FastAPI
RAG
LangChain
MS SQL
Docker
K8s
OCR
2022
Jan 2022 — Feb 2025

End-to-End Developer & Data Scientist

Fiverr & Upwork

Remote — US, UK, EU Freelance

100%

Client Score

40+

Projects Delivered

5★

Top Rated

I delivered ML pipelines, React/Next.js dashboards, automation systems, and data science solutions to international clients. I worked across a broad range — from a deep-learning CAPTCHA-solving model to distributed scraping infrastructure.

End-to-end ML pipeline: data collection → model training → API deployment
Client-specific dashboard and portal apps built with React / Next.js
Deep-learning-based CAPTCHA-solving model: CNN trained on a 200×50 PNG dataset
Distributed web scraping infrastructure: Selenium + aiohttp, 100K+ pages per hour
XGBoost churn-prediction model: 91% AUC score
Python automation saving clients 20+ hours per week
100% client satisfaction — Top Rated 5★ profile on every platform
Python
React
Next.js
TypeScript
Node.js
Scikit-learn
XGBoost
PostgreSQL
REST API
Selenium

Academic

Education.

Istanbul University

Economics

Bachelor's

StatisticsA/B TestingEconomicsData Analysis

Istanbul Gelisim University

Computer Engineering

Bachelor's

AlgorithmsAgileSoftware EngineeringData Structures

Contact

Let'swork together.

Got a project in mind, or thinking about a collaboration? Let's talk.