I am Niko
About Me
I am an informatics student with hands-on experience in data-driven engineering, analytical modeling, and applied software development. My work combines software engineering, data analysis, and algorithmic thinking to solve complex technical problems and build practical, reliable solutions.
I work across the full lifecycle of technical systems: from data processing, mathematical modeling, and optimization to backend development, automation, and user-facing tools. My experience includes geodata and mobility analytics, transport modeling with GTFS data, time-dependent routing, computer vision pipelines, and applied machine learning.
With a background spanning both engineering and computer science, I bring a structured and research-oriented approach to every project. I focus on turning messy data and difficult requirements into clear models, efficient systems, and measurable results.
Experience & Education
Education
B.Sc. Informatics
Broadly focused on software engineering, algorithms, full-stack web development, data science, artificial intelligence, cloud computing, and geoinformatics, with strong practical emphasis on applied software projects and analytical problem solving.
- ▸Advanced coursework in Algorithms, Advanced Algorithms, and Data Structures
- ▸Strong foundation in Software Engineering, Software Architecture, and Software System Development
- ▸Applied focus in Full Stack Web Development, Data Science, Cloud Computing, and Computer Networks
- ▸Additional specialization in Applied Artificial Intelligence, Multimodal Artificial Intelligence, Applied Robotics, and Geographical Information Systems
M.Sc. Petroleum Geology
Specialized in mathematical modeling, statistical analysis, uncertainty assessment, and data-driven pattern recognition applied to geological and reservoir systems.
- ▸Mathematical modeling for geological and reservoir-related analysis
- ▸Monte Carlo simulations for uncertainty analysis
- ▸Pattern recognition in large geological datasets
Work Experience
Student Research Assistant (HiWi)
Supporting research and applied analytical work in geodata and informatics, with a focus on mathematical modeling, spatial data analysis, transport analytics, and algorithmic problem solving.
- ▸Built mathematical models for estimating population figures and evaluating the popularity of transport hubs
- ▸Worked with GTFS-based transport data to model mobility patterns and implemented time-dependent routing logic, including Dijkstra-based travel time calculation
- ▸Contributed to geospatial analysis, statistical evaluation, and Python-based analytical workflows
Project Engineer
Worked in an engineering role with a strong focus on data analysis, statistics, automation, and predictive modeling for geological and reservoir-related tasks.
- ▸Automated complex database workflows across hundreds of tables using Python and VBA
- ▸Improved processing efficiency by replacing manual data workflows with automated pipelines
- ▸Built predictive and analytical models for engineering decision support
Skills
Software Engineering
- • Python
- • Java
- • С and C++
- • JavaScript
- • API Integration
Data Science & Analytics
- • SQL
- • Statistical Analysis
- • Data Validation
- • Forecasting
- • Data Processing
AI & Machine Learning
- • Computer Vision
- • YOLO
- • Multimodal AI
- • VLM Fine-Tuning
- • Dataset Generation
- • Model Evaluation
Algorithms & Optimization
- • Graph Algorithms
- • Dijkstra
- • Integer Linear Programming
- • Constraint Modeling
- • Performance Optimization
- • Caching
Automation, DevOps & Engineering
- • Process Automation
- • Data Pipelines
- • Docker
- • Git
- • Containerization
- • Deployment
GeoData & Mobility Analytics
- • Geodata Analysis
- • Geodata Modeling
- • Mobility Analytics
- • Transport Data
- • Spatial Thinking
- • Algorithmic Analysis
Projects
Digital Board Game
Star Pilots
Improved and refactored version of the original university project: a cooperative two-player board game implemented in Java and LibGDX. Reworked into a layered architecture (domain/application/ui) with Command + State Machine + Observer patterns, JSON-driven campaigns, centralized resource management, and browser deployment via TeaVM with Docker/Nginx hosting.
Full-Stack Fitness Platform
Team-A Training Club
End-to-end development of a comprehensive platform from UI/UX design to backend implementation. Built with Laravel for robust server-side logic and React for a dynamic, responsive user interface, containerized with Docker. Features include advanced user management, role-based access control, and real-time data synchronization.
Kubernetes-Deployed Appointment Service
TERMINator
Cloud-native web application built with containerized architecture for high availability and horizontal scaling. Leverages Docker for consistent deployment environments and Kubernetes for orchestration, auto-scaling, and fail-safety. Implements microservices architecture with service mesh for resilient inter-service communication.
Public Transit Routing Engine
GTFS Time-Dependent Pathfinder
Python tool to compute shortest paths on public transit networks from GTFS feeds using time-dependent Dijkstra that respects real timetables (departures/arrivals/transfers) including overnight times (>24:00:00). Implements GTFS loading with Parquet caching, builds a time-dependent graph, supports forward (earliest arrival) and backward (latest departure) variants, and includes unit/integration tests.
Multimodal AI Fine-Tuning Pipeline
Multimodal AI for Checkers
End-to-end pipeline to fine-tune a Vision-Language Model to understand and play International Checkers from board screenshots. Includes bot-vs-bot gameplay for data generation, structured VQA dataset creation, LoRA fine-tuning of Qwen2-VL-2B, and evaluation comparing pre- vs post-training; exposes interaction via a Flask API.
Autonomous Person Detection Backend Pipeline
DJI Drone SAR/Security System
Server-side architecture for processing live camera data from a DJI drone system with real-time person detection (YOLO). Built a complete backend pipeline plus Telegram/Discord bots, alerting, and a Streamlit web interface for monitoring and alarms
Contact
Send a Message
Get in Touch
If you would like to collaborate or discuss an opportunity, reach me via email or LinkedIn
© 2026 Nikolai Krysin
Designed and built by me and my wife