Ankara, Türkiye
Senior AI Systems Engineer & Full-Stack Architect with 20+ years of experience delivering scalable, secure, and business-aligned software solutions across multiple industries including SaaS, fintech, healthcare, logistics, media streaming, and renewable energy. Proven expertise in end-to-end software development lifecycle (SDLC), from architecture design and backend systems to frontend applications, mobile platforms, and cloud-native deployments. Strong background in building cost-effective, high-performance systems with a focus on security, scalability, and operational efficiency, combined with the ability to work independently and drive complex projects with minimal supervision. Extensive experience in developing B2B/B2C SaaS platforms, booking engines, payment integrations, tourism and property portals, medical and healthcare systems, live streaming infrastructures (RTMP, RTSP, SRT), CDN and IoT-based solutions, logistics and automotive systems, as well as advanced engineering projects in hydrogen electrolyser and fuel cell technologies. Deep expertise in enterprise architecture, cybersecurity, and cloud security, including penetration testing, threat modeling, vulnerability assessments, and secure system design. Experienced in architectural reviews, solution design, POCs, pre-sales engineering, and full-cycle deployment of enterprise-grade systems. Currently focused on AI-driven systems and intelligent automation, designing and building advanced platforms using Large Language Models (GPT, Claude, Mistral), Retrieval-Augmented Generation (RAG), and multi-agent architectures. Strong hands-on experience with Python (AI/ML), Node.js/NestJS, React/Next.js, and modern data engineering stacks including Airflow, dbt, Pandas, and Spark, combined with cloud infrastructure (AWS, Azure, Kubernetes, Docker), CI/CD pipelines, and scalable API architectures (REST, GraphQL). Actively applying AI technologies to complex and regulated domains such as biopharmaceutical manufacturing and advanced cell therapy systems (CAR-T / ATMP). Designing data-driven, intelligent pipelines for optimizing biological workflows including T-cell processing, integrating machine learning with laboratory systems for real-time monitoring, predictive analytics, anomaly detection, and automated quality control within GMP-compliant environments. Focused on building next-generation AI-powered decision systems, digital twins, and autonomous platforms that improve efficiency, reproducibility, and outcomes across both industrial and scientific domains.
BioMind / CellAI / ImmunoAI – AI-Supported Cellular & Molecular Health Platform BioMind (CellAI / ImmunoAI) is an AI-driven healthcare platform I am currently developing, focused on combining cellular biology, molecular diagnostics, and intelligent software systems into a unified architecture. The system integrates laboratory-level processes such as PBMC isolation, cell culture, flow cytometry, apoptosis analysis, DNA/RNA extraction, and PCR/qPCR workflows with an AI-based data processing layer. These biological signals are transformed into structured datasets and analyzed using advanced AI models to generate interpretable outputs for clinical and research use. A key focus area of the platform is immune system profiling and cellular behavior analysis, enabling deeper insights into patient-specific biological responses. In parallel, the system is designed to support emerging therapeutic domains such as CAR-T cell therapies and cell engineering workflows, where cellular data, gene expression patterns, and treatment responses can be analyzed and optimized through AI. From a technical perspective, the platform is built around: - Data pipelines that convert wet-lab outputs into machine-readable formats - AI/LLM-based analysis layers (including RAG and multi-agent systems) - Scalable backend services and modular APIs - Interfaces for clinicians, laboratories, and research teams The goal is to create a computational layer on top of biological systems, enabling: - Early signal detection in complex biological data - Clinical decision support systems - Standardized analysis of cellular and molecular processes - Integration of lab workflows with digital health platforms This project represents a convergence of biotechnology, bioinformatics, and AI engineering, with the long-term vision of building a scalable, AI-powered infrastructure for precision medicine and advanced therapies.
AI-Enabled Health Technology Platform (Oncology & Cell-Based Systems) Designing and developing a health technology platform that integrates artificial intelligence, advanced software systems, and biomedical infrastructure, with primary application in oncology and cell-based research environments. The project is structured as a platform-level system enabling the digital management, traceability, and integration of data, devices, and operational workflows used in complex biomedical settings. Artificial intelligence is implemented as a decision-support and enablement layer, assisting with: • Structured processing of biomedical and clinical knowledge • Modeling and organization of multi-source biological and operational data • Analytical workflows supporting complex evaluation and comparison scenarios • Traceable, human-in-the-loop decision processes The software architecture is designed to be modular, secure, and regulation-aware, incorporating: • LLM-based knowledge processing systems • Retrieval-augmented architectures for integrating scientific literature and protocols • Role-based access and workflow orchestration • Auditable data pipelines aligned with compliance and quality requirements The platform integrates with laboratory and processing hardware, providing end-to-end visibility across data, devices, and workflows, and is actively maintained through continuous development and technical validation. Collaboration opportunities are open with technology teams, academic institutions, and biotechnology infrastructure partners.
AI-Powered E-Commerce Infrastructure with LLM, RAG, MCP, NLP & Multi-Agent Orchestration Designed and implemented an end-to-end AI-driven e-commerce platform combining modern web/mobile technologies with advanced AI orchestration. The system enables customers and admins to interact via intelligent, natural language interfaces powered by LLM, RAG, MCP, and multi-agent pipelines. Key Features: • AI Chat: Human-like shopping assistant using LLMs (Mistral, LLaMA, GPT, Claude, Gemini). • RAG Pipeline: Search/retrieval across millions of records (PDF/Excel) via VectorDB (Chroma, Pinecone, pgvector). • MCP Integrations: Real-time APIs, CRM, email, Slack, WhatsApp. • Multi-Agent Orchestration: LangChain + LlamaIndex coordinate agents (product, supplier, crawler, support). • Web Crawlers: Automated data from 100+ e-commerce sites (Playwright, Puppeteer, Scrapy). • NLP Modules: Hugging Face & spaCy for categorization, translation, sentiment. • Automation: n8n workflows for notifications, stock alerts, CRM sync. • Omnichannel Experience: – Frontend Portal: Next.js + Tailwind + SEO. – Admin Dashboard: Next.js + Tailwind Admin Template. – Mobile App: React Native + Nativewind. Infrastructure: Self-hosted Ollama for local LLM inference, Docker/Kubernetes for deployment, PostgreSQL core DB, Nginx proxy, microservice-ready architecture. Technologies: Backend: NestJS, Node.js, LangChain.js, LlamaIndex.js Frontend: Next.js, TailwindCSS, SEO Admin: Next.js, Tailwind Admin Template Mobile: React Native, Nativewind LLM: Mistral, LLaMA, GPT, Claude, Gemini RAG / VectorDB: Chroma, Pinecone, Weaviate, pgvector NLP: Hugging Face, spaCy Integrations: MCP, n8n Crawler: Playwright, Puppeteer, Scrapy Database: PostgreSQL Infra: Docker, Kubernetes, Nginx
Intelligent SCADA & Monitoring Platform for Hydrogen Electrolysers & Fuel Cells We are developing a fully custom industrial control and monitoring software platform from scratch. The system will manage AEM/PEM Hydrogen Electrolysers and Fuel Cells in real-time, with a strong focus on data visualization, intelligent decision-making, and system efficiency. The project includes: * A modern SCADA system for real-time control, monitoring, and alarm/event handling * AI-based modules for predictive maintenance, anomaly detection, and process optimization * Web-based dashboard for real-time data display and operator control * Secure remote access, user role management, and scalable architecture * Future integration of LLMs (Large Language Models) and Manufacturing Control Platform (MCP) Server. Technology Stack & Tools We’ll Use in This Project: Core Technologies: * Industrial Communication: MODBUS, OPC UA, or MQTT (to interface with electrolysers and fuel cells) * Backend: Python (FastAPI) or Node.js for APIs and control logic * Frontend: React.js for operator dashboard and real-time data visualization * Databases: PostgreSQL (for metadata, user control) + InfluxDB or TimescaleDB (for sensor/time-series data) * Real-Time Communication: WebSocket or MQTT for telemetry and control * Deployment Architecture: Docker + Microservices AI/ML Capabilities (Later Stages): * Predictive Maintenance with scikit-learn, XGBoost, or other ML frameworks * Anomaly Detection using Autoencoders, Isolation Forests, etc. * Time-series forecasting and pattern recognition * Reinforcement learning for optimizing hydrogen production parameters * Visual analytics through Grafana or integrated custom charting LLM Integration (Future): Use of Large Language Models (e.g., OpenAI GPT, LLaMA) for: * AI-based assistant for interpreting alarms, reports, and system insights * Enhanced multi-lingual support for operator environments * Integration with internal logs and data (RAG pipelines)
I recently designed and developed a complete AI assistant solution for customer support that enhances both pre-sales and post-sales interactions. The system is powered by Large Language Models (LLMs) and fully integrated with SAP, Salesforce, and our e-commerce platform. Key Features and Capabilities • Conversational AI chatbot built with OpenAI GPT-4 and LangChain • Retrieval-Augmented Generation (RAG) using ChromaDB to deliver accurate, context-aware answers based on internal documentation, product manuals, and knowledge base content • Real-time SAP integration to retrieve order status, customer data, and inventory information • Salesforce CRM integration to manage leads, support tickets, and customer interactions • E-commerce integration to provide product details, cross-sell suggestions, and handle shopping-related queries • Multilingual support, sentiment analysis, and automatic escalation to human agents when needed • Secure API connections with OAuth2, JWT, and role-based access control • Dashboard and analytics to monitor assistant usage, response accuracy, customer satisfaction, and ticket resolution rates ⸻ Technology Stack • LLMs: OpenAI GPT-4 (Claude 3 tested) • AI Frameworks: LangChain, FastAPI • Vector Database: ChromaDB (Pinecone was also tested) • System Integrations: SAP BTP via ODATA API, Salesforce REST API, WooCommerce API • Backend: Python with FastAPI • Deployment: Docker, Azure App Services • Frontend Integration: Custom web widget and embedded panels in CRM systems