I build applied AI systems that connect models, software, data, and edge hardware. My portfolio is organized around three career tracks: AI Engineering, Economics & Financial Technology, and Embedded Software Engineering.
LinkedIn · Medium · YouTube · Email
Computer vision, edge AI, audio AI, LLM applications, Azure Machine Learning, YOLO, ONNX, TensorFlow Lite, RTSP analytics, and production-oriented evaluation.
View the AI Engineering portfolio
Economic data platforms, financial forecasting, risk analytics, fraud detection, portfolio research, and reliable backend systems for financial applications.
View the Economics & FinTech portfolio
C++, Rust, Raspberry Pi, ESP32, embedded Linux, device provisioning, sensor systems, industrial protocols, and edge-to-cloud integration.
View the Embedded Software portfolio
| Project | Area | What it demonstrates | Status |
|---|---|---|---|
| zone-analysis | AI Engineering | C++, OpenCV, YOLO ONNX, RTSP ingestion, polygon zones, dwell-time events | In Progress |
| hailoai | AI Engineering / Embedded | Raspberry Pi 5 and Hailo NPU edge-inference benchmarking | In Progress |
| baby-cry-ai-detector | AI Engineering | Audio classification, source-level evaluation, threshold selection, honest error analysis | In Progress · Private |
| MacroTR | Economics & FinTech | TCMB EVDS ingestion, FastAPI, PostgreSQL, React, economic data visualization | In Progress |
| Finstack-Starter-Lite | Economics & FinTech | FastAPI, React, PostgreSQL, Redis, Docker, tests, and CI for financial applications | Completed v0.1 · Private |
| edge-wifi-provisioning-showcase | Embedded Software | Headless device onboarding, AP-to-STA flow, rollback, and security-conscious documentation | Completed |
| skyvision-control-case-study | Embedded / AI Systems | Sanitized ESP32, telemetry, Flutter, RTSP, and edge-vision system case study | In Progress |
- Make edge-AI projects reproducible with benchmarks, architecture diagrams, tests, and privacy-safe demos.
- Build finance projects around time-aware validation, interpretable baselines, and clear limitations.
- Add embedded evidence through protocol implementations, hardware-in-the-loop tests, and documented device constraints.
Languages: Python, C++, Rust, SQL, Dart, TypeScript
AI & Data: PyTorch, TensorFlow, scikit-learn, OpenCV, YOLO, ONNX, TensorFlow Lite, Azure ML
Backend & Platform: FastAPI, PostgreSQL, Redis, Docker, GitHub Actions
Edge & Embedded: Raspberry Pi, Hailo AI HAT+, ESP32, embedded Linux, RTSP
Frontend: React, Flutter
- Medium: medium.com/@sirvanksc
- Planned series: edge AI benchmarks, trustworthy financial ML, and practical embedded systems
- Personal portfolio: add verified URL
- LinkedIn: linkedin.com/in/srvnksc
- Medium: medium.com/@sirvanksc
- YouTube: youtube.com/@SirvanKesici
- GitHub: github.com/sirVnK
For engineering roles, project discussions, and technical collaboration:
- Email: sirvan@sophtrun.com.tr
- LinkedIn: Şirvan Kesici
- YouTube: @SirvanKesici
Repository status labels describe the current portfolio state. Planned work is listed as planned and is not presented as completed experience.