🔗
   Senior Software Architect / Engineer / Full-Stack Developer position, bringing 28 years of hands-on experience across a wide range of programming languages — first of all C++ — plus the tools and environments around them, backed by strong DevOps skills. Passionate about clean, best-practice engineering in computer vision, Web, IoT (Internet of Things), UI/UX, distributed systems and synchronization, and AI (neural networks & LLMs).

   What sets me apart: I don't just close the ticket - I solve the real problem behind it. Hand me a small task and, if I spot a deeper issue underneath, I won't stop at the requested patch; I dig in and fix the whole thing properly. I run comprehensive analysis of both the solution and the project as a whole, and I own the entire stack - architecture, backend, frontend, DevOps, and infrastructure - not just code, which is unusual for a single developer. And I don't just shave a few milliseconds off a function - I re-architect the whole approach: where a team of ~20 engineers had spent 2 years getting a real-time video pipeline down to ~1.5 seconds, I re-designed it around many tightly synchronized threads and reached ~6 milliseconds in about 2 months - squeezing performance out of the very same hardware that most engineers wouldn't believe possible. The goal is always the same: flawless, production-ready software, delivered in the shortest possible time.

🔗

Programming Languages

   • C/C++ (expert, 28 years) - GCC, Clang, MSVC, STL, ANSI C, MFC, ATL, strong OOP
   • Python, Bash, PHP, Perl, Java, C#, JavaScript/Node.js, HTML/CSS/REST
   • SQL/T-SQL, CUDA/OpenCL (parallel computing on GPUs)
   • Embedded C/C++ (ESP32/ESP8266/Arduino/STM32)

Frameworks & Tools

   • STL, Boost, MFC, ATL, Qt, .NET, ASP.NET
   • GCC, Clang, MSVC, GDB, Visual Studio, VS Code, Qt Creator, Eclipse, Arduino Studio
   • Git/GitLab/GITEA/Azure DevOps, SVN, VSS, Source OffSite
   • REST/JSON, YAML, WebSockets, OpenMP
   • Installers: WIX, NSIS, InstallShield, Wise, Self-Deployable stubs

Platforms

   • Windows (desktop, embedded, server)
   • Linux (Debian, Ubuntu, Raspbian, OpenWRT) - development + server administration
   • Cloud: AWS (Bedrock AI, EC2, S3), Azure (VMs, Storage, DevOps)
   • IoT: Embedded/RTOS/VxWorks, ESP/STM, Raspberry Pi
   • Mobile: Android, Windows Mobile, Palm OS

DevOps & Infrastructure

   • CI/CD, Bash/Python automation, DevOps scripting
   • Cloud infrastructure (AWS/Azure), load balancing, monitoring, performance tuning
   • Docker containerization, Kubernetes orchestration
   • Linux server administration: Nginx/Apache, Exim (email), MySQL, WireGuard/OpenVPN (VPN)
   • IPtables (firewall), fail2ban (intrusion prevention)

Databases

   • MySQL, PostgreSQL, Microsoft SQL Server, SQLite, NoSQL, FastObjects/POET
   • DB design, query optimization, performance tuning, migration, embedded
   • Experience in regulated medical device environments

Specialized Domains

   • Computer vision - OpenCV, DNN (Deep Neural Networks), object/face recognition/detection/tracking
   • AI/ML - LLMs (AWS Bedrock, GPT4All), neural networks (CUDA-accelerated)
   • Big Data & distributed computing - multi-server networks, Kubernetes orchestration
   • Web development - full product lifecycle, payment processing, API, AI chatbots
   • UI/UX - custom GUI frameworks, rich text editors with tables, skinned interfaces
   • Localization & internationalization - UTF-8/UTF-16, multi-byte, Asian/RTL languages
   • Medical device software (IEC 62304, FDA-regulated environments, embedded systems)
   • IoT systems - mesh networks, LoRa, GPIO/PWM/ADC/DAC, I2C/SPI/UART/CAN protocols
   • Blockchain - consensus algorithms, real-time trading, wallet-to-wallet payments

Technical Highlights

   • Multi-threading, inter-process communications (Linux, Windows, CUDA, IoT)
   • Custom C++ memory management (leaks detection, garbage collection, caching)
   • Advanced printing support (custom dialogs, WYSIWYG preview, metafile-based engine)
   • Compression algorithms (custom LZW/Huffman for embedded, ~5KB memory footprint)
   • Encryption, cybersecurity
   • Strong debugging skills, memory leak detection, real-time execution logging
   • Code optimization for resource-constrained environments

Soft Skills

   • Self-organized, reliable, strong ownership, bias for action
   • Team leadership, cross-department coordination (QA, tech support, sales, marketing)
   • Agile/Scrum/Kanban, code reviewing
   • Long-term remote/hybrid experience (since 1998), multiple time zones
   • Ability to work under extreme conditions (infrastructure damage, blackouts during war)
   • AI-assisted productivity tools (Cursor, Claude, ChatGPT, Copilot) - adhering to company policies

Languages

   • English (fluent)
   • Ukrainian (native)
   • Russian (near-native)
   • German, Polish, French (limited working proficiency)

🔗
   28 years of experience in C/C++ and many other languages, for a wide range of platforms, including, but not limited to: GCC/Clang/GDB, STL, ANSI C, MFC, ATL, strong OOP skills, CUDA/OpenCL, Graphical User Interface skills, and a broad range of appliances integrated with each other — from huge BIG-DATA multi-server application networks (analysis/identification/recognition/linking/categorization of massive amounts of information) down to tiny IoT devices on Arduino/ESP32/ESP8266 controllers and modules, and Palm OS (both application development and synchronization with desktop — the platform is now obsolete, but we were first on the market, so it is real-world experience building tiny, efficient mobile apps on very limited resources), Internet protocols, localization of applications, web-development (PHP/ASP.NET/C#/Node.js/JavaScript/HTML/REST/JSON/YAML), medical device software and DB development, administration of Linux-based servers (web-sites, email, VPN, etc).

   Platforms: Linux (POSTFIX, coding in C/C++, Bash, Python, PHP, Node.js, JavaScript, Perl, administration of cloud servers/hosts/containers and web-sites, security of sites/hosts, load-balancing, automated administration of virtual cloud servers), Windows (C++, GUI, ASP.NET, C#), Windows Mobile (C++ coding for Smartphone editions), Android, Palm OS (developing mobile apps since 1998), DOS, RTOS, Embedded systems (Raspberry PI), smart IoT controllers/devices (on ESP32/ESP8266/Arduino).

   Programming languages and instruments. 28 years of continuous daily C/C++ development (GCC, Clang, MS Visual C++, MS Visual Studio Code, Qt Creator, Eclipse, Metrowerks CodeWarrior C++, Arduino Studio C++). Long-standing experience in a wide range of programming languages: PHP, Bash, Python, Perl, Java, ASP C#, Node.js, JavaScript (standalone and within HTML), HTML/JS/CSS, and help-file editors. Installer creation (WIX/MSI, Nullsoft Install System, InstallShield, Wise), sound editors (Adobe Audition, FL Studio), graphics design (e.g. Adobe Photoshop, Illustrator, GIMP, CorelDRAW & Photo-Paint), video editors (CapCut, Pinnacle), and diagram design. Work with version control systems and code review in Git/GitLab/Gitea/Azure DevOps/SVN/VSS and Source OffSite. DB development for MySQL/NoSQL/Microsoft SQL/local SQLite/FastObjects - DB design, query optimization, performance tuning/monitoring/troubleshooting, legacy DB migration and abstraction layer design, compliance with regulated medical device environments.

   Other experience: Self-organized, team player, bias for action, deliver results, remote work/hybrid experience (since 1998) in different time zones, coordinating work of various departments (QA, tech support service, sales, marketing) using Agile/Scrum/Kanban/CI/CD methods, experienced in writing highly optimized (and well readable), stable code, blockchain technology (ex: blockchain algos, real-time trading, sales in direct wallet-to-wallet payments). Self-organized, able to work under stress and heavy conditions when infrastructure is damaged - have real experience working during a war: through near-daily air-raid alerts (while russia bombed our cities throughout 2022–2023 and destroyed a large part of Ukraine's critical infrastructure) I kept work running under continuous blackouts and loss of landline Internet by building an off-grid solar power station with a large power reserve, able to switch to mobile Internet automatically whenever the main fiber line went down. Localization of Software/sites to multiple languages, creation of installers and smart download stubs.

🔗
   - Distributed computing for Big-Data coordinated dynamically allocated network of Kubernetes/Docker + web-based control panel to configure/monitor such system. Scaleable VPS cloud Kubernetes (nodes/containers) cloud (AWS, Azure, etc.) allocation to adjust to demand automatically.
   - Cloud platforms and containerization: AWS, Azure, Docker containerization for development and production deployments, Kubernetes orchestration.
   - Multi-threading, inter-process communications. In Linux, Windows, and IoT devices. Have experience in OpenMP parallel code execution, and, OpenCL and CUDA parallel computing on GPUs.
   - Strong debugging skills, memory-leak detection techniques, and real-time execution logging. Have experience debugging on IoT devices with a very limited set of features (no screen, no disk, no Internet, etc.).
   - Computer vision: OpenCV, object detection/identification/tracking, Deep Neural Networks (DNN). Experienced in running various neural-network models linked to OpenCV/DNN on hardware with limited resources, as well as on systems with powerful video accelerators (CUDA-based neural networks), implemented in both Python and C++.
   - Custom C++ memory management experience (automatic leaks detection, garbage collection, faster operations on small memory blocks, caching, sharing, statistics on memory usage, etc).
   - Advanced printing support - built a complete printing C++ library, from a custom print dialog all the way to WYSIWYG print preview (powered by a very fast metafile-based graphics engine).
   - Experience of work with localized strings in UTF-8/UTF-16, multi-byte and unicode formats, can deal with Asian and right-to-left languages (Hebrew, Arabic), know possible issues that may arise there.
   - Knowledge of compression algorithms. Developed a custom implementation of LZW and Huffman encoding for mobile/embedded systems that needs almost no memory (max ~5 KB) for packing/unpacking, whereas all standard implementations have a huge memory footprint.
   - Built an ergonomic UI library and have experience coding a full set of fully custom user-interface objects for Windows - in other words, visual and audio themes. My most complex achievement here was a fully featured rich-text editor with tables support.
   - Created graphical library (Windows GDI-based) which supports fast double-buffering, secure management on all kinds of pen/brush/font objects (which is typically the source of bugs in Windows programs, in this library work with such objects is absolutely safe), also it includes set of very fast classes for direct work with bitmaps (quick pixels access, color adjustment, filtering, etc).
   - Code style - doing my best to be perfect in it, always keeping it clear (fine naming is always the case), well structured, commented, secure, easy to read.
   - Web-sites development. Full product life-cycle, payment processing, marketing, newsletters, API, AI LLM chatbots, etc. Using Python, PHP, ASP.NET, C#, Node.js, JavaScript and Perl, plus CSS, JSON, YAML and XML, and creation of slim, highly optimized web graphics in Photoshop and Illustrator. Have experience coding Angular frontends. Can even build a whole web server in C++ or Python - such servers are incredibly fast, need very few resources, and are highly resistant to DDoS attacks and other cybersecurity threats.
   - Linux/Debian experience - administration and programming/scripting. Hosting Nginx/Apache web servers. Exim mail server administration, MySQL database administration, building private VPN networks (WireGuard, OpenVPN), firewall (IPtables), fail2ban intrusion-prevention system. Can create bash scripts/images for DevOps to quickly deploy systems with needed configurations. Debian, Ubuntu, OpenWRT, Raspbian.
   - Internet of Things (IoT) experience, including Raspberry Pi and a number of ESP32/ESP8266/STM/ARM controllers with various modules, all coordinated via a Raspberry Pi Linux/Debian/Raspbian system. Knowledge of GPIO, PWM and ADC/DAC signal generation and processing, I2C, SPI, UART, custom PWM-based and CAN protocols, mesh networks and LoRa communication.
   - Embedded OS - know how to write small, fast code on very limited resources. Have experience with RTOS, including asynchronous multi-threading support and autonomous systems.
   - Have 4 years of experience with Delphi/Pascal and linked to it Borland/Embarcadero C++ Builder implementation.
   - Blockchain/Cryptocurrency. Deep knowledge of blockchain techniques, consensus algorithms. Implemented automated Bitcoin store checkout page for immediate direct purchases. Implemented a real-time trading tool that makes instant AI decisions and trades automatically via the Poloniex and Bittrex platform APIs.
   - Medical device software/DB architecture: conducted full architectural analysis and performance profiling of a proprietary embedded database system across a fleet of 17,000+ clinical instruments; produced IEC 62304-compliant technical documentation.
   - I leverage AI-powered productivity tools in my coding practice to boost efficiency and creativity - among them Cursor, Claude, ChatGPT, and GitHub Copilot. I always adhere to each company's policies and guidelines on the use of such tools.

🔗

🧑‍💼 Senior Software Engineer (C++, Real-Time Computer Vision, High-Throughput Pipelines)

🏢 Walmart

📅 May 2026 - now (2 months)

Project description:
   High-throughput barcode-scanner stream-processing pipeline for a nationwide retail fleet (targeting rollout across 12,000+ stores).

Responsibilities:
   - Re-architected a real-time barcode recognition pipeline that a team of ~20 engineers had spent ~2 years optimizing down to ~1.5 s, and reached ~6 ms within my first ~2 months on the project - measured on the existing production codebase.
   - Designed a non-blocking O(1) scanner API: heavy neural recognition runs on dedicated worker threads while API calls return in constant time under a minimal-hold-time lock, serving the latest fully processed frame. Per-frame freshness is deliberately traded for guaranteed zero-wait response - the right call for high-throughput scanning at any call rate.
   - Built the result hand-off to be race-free by returning a full deep copy of the detected barcode array to each caller, eliminating shared-state lifetime and tearing issues entirely.
   - Focused on squeezing maximum throughput out of existing hardware through tightly synchronized multi-threaded processing, cache-friendly data flow, and elimination of redundant work on the hot path.

   🔧 Tools: C++ (STL, multithreading), real-time computer vision, neural inference, OpenCV, high-throughput pipeline design.

🔗

🧑‍💼 Senior Software Engineer (C++/C#, Database Architecture, Medical Devices)

🏢 BD (Becton, Dickinson and Company. NYSE: BDX).

📅 Nov 2025 - Feb 2026 (3 months)

Project description:
   Embedded medical software. Database Architecture.

Responsibilities:
   - Conducted comprehensive analysis of the existing proprietary database architecture of blood culture system (17,000+ instruments worldwide); produced a detailed technical report covering data distribution, storage patterns, object lifecycle, and performance profiling of bottlenecks.
   - Designed and implemented a Proof of Concept for a drop-in C++ and C# database abstraction layer replacing a commercial proprietary engine (FastObjects); POC is fully operational on a representative subset of the codebase and ready for production rollout across the entire fleet.
   - Conducted 7-vendor platform evaluation and produced technology selection rationale - adopted by Software Engineering, Systems Engineering, QA, and Cybersecurity stakeholders - eliminating millions of dollars in annual licensing costs and reducing overall instrument cost of goods.

   🔧 Tools: Visual C++, C#, Microsoft SQL Server, ADO.NET, FastObjects/POET, Windows Embedded, VxWorks.

🔗

🧑‍💼 Senior Software Engineer (C++ and Neural Networks for computer vision)

🏢 Ragnarok Group / Spiral Networks (SpiAlert).

📅 July 2024 - Nov 2025 (1 year 4 months)

Project description:
   Enterprise AI Cybersecurity & Sensitive Data Protection Platform
   Designed and developed an AI-powered computer vision cybersecurity platform for protection of sensitive corporate data and restricted workstations. The system continuously monitored users and surrounding environment through webcams and used real-time behavioral and scene analysis to automatically detect unauthorized observation attempts, suspicious presence, privacy threats, and potential data leakage situations. When a workstation entered a dangerous state — for example presence of an unauthorized person behind the legitimate user, suspicious gaze direction, or attempted observation/recording of confidential information — the platform could automatically lock workstation access, dim or disable displays, hide sensitive information, and trigger security response workflows in real time using advanced computer vision and AI analysis.

Responsibilities:
Client-Side AI Engine (C++ / Real-Time Computer Vision / Cybersecurity)
   - Architected and implemented large-scale real-time AI-powered computer vision platform focused on protection of sensitive corporate data, behavioral analysis, scene understanding, and autonomous cybersecurity threat detection directly on protected workstations.
   - Designed the overall client-side AI processing architecture including multi-stage inference pipeline, asynchronous scene analysis, object tracking, adaptive resource balancing, low-latency frame processing, and high-efficiency multi-threaded video analysis optimized for continuous enterprise operation.
   - Conducted extensive R&D (Research & Development) activities focused on identifying optimal architectural, AI, and performance strategies for real-time cybersecurity-oriented computer vision systems operating under strict hardware and latency constraints.
   - Designed ultra-efficient multi-threaded video processing architecture utilizing up to 7 highly synchronized worker threads which maximized modern multi-core CPU utilization through asynchronous staged frame analysis and pipeline parallelism.
   - Engineered real-time concurrent frame-processing strategy where different pipeline stages operated simultaneously on different temporal states of the scene. While one thread performed fresh-frame motion analysis and ROI recalculation, other threads continued risk assessment, object tracking, neural inference, and behavioral analysis based on previously stabilized scene state, dramatically improving throughput and reducing end-to-end latency.
   - Built zero-copy / near-zero-copy frame processing pipeline with extreme focus on memory efficiency, cache locality, and elimination of redundant image operations. The system minimized unnecessary frame cloning, intermediate buffers, color-space conversions, and additional processing layers frequently present in traditional computer vision stacks.
   - In optimized hardware scenarios the pipeline operated directly on DMA-backed webcam frame buffers, processing original camera frames through pointer-based access without redundant memory copies. This enabled exceptionally high throughput and extremely low CPU overhead even during continuous real-time AI inference.
   - Developed intelligent scene-analysis and motion-detection subsystem which dynamically minimized neural-network processing regions (ROI - Region of Interest) and adaptively tuned inference frequency, frame rate, and scene refresh logic depending on scene activity and cybersecurity risk level.
   - Implemented adaptive performance optimization algorithms which dynamically balanced frame-rate, inference cadence, ROI recalculation frequency, scene stabilization, and motion sensitivity in real time to maintain stable autonomous operation on low-end hardware with minimal processor and memory consumption.
   - Achieved extremely low hardware utilization allowing the ML engine to operate continuously even on low-end CPUs while still performing real-time analytics, object tracking, scene correlation, behavioral analysis, and cybersecurity threat detection.
   - Developed persistent multi-object and multi-person tracking subsystem capable of correlating faces, bodies, objects, movement vectors, behavioral anomalies, and risk zones across video streams with temporal scene consistency and live analytics.
   - Implemented resilient offline-first telemetry and event synchronization architecture allowing protected client workstations to continue autonomous operation during temporary network outages. Security-critical events, behavioral analytics, risk detections, and monitoring metadata were securely buffered in protected local storage and automatically synchronized with centralized infrastructure once connectivity was restored.
   - Designed and implemented secure local protected storage subsystem for temporary buffering and persistence of sensitive cybersecurity telemetry, AI detection events, behavioral analytics, snapshots, and monitoring metadata before secure synchronization with centralized infrastructure. Implemented additional protection mechanisms for locally stored sensitive data including encrypted storage, integrity validation, secure buffering, and restricted-access handling to reduce risks of data leakage or forensic extraction from protected endpoints.

Machine Learning / AI / Neural Networks
   - Developed advanced hybrid neural-network based computer vision system for real-time detection, identification, and behavioral analysis of people and objects in monitored environments.
   - Implemented AI-driven privacy and threat analysis modules capable of identifying unauthorized persons, detecting mobile phones in hands, estimating camera orientation, and evaluating whether sensitive or protected information displayed on monitored screens could potentially be photographed, recorded, or leaked.
   - Worked extensively with YOLO-based object detection models for real-time human/object detection, scene analysis, behavioral event tracking, and threat identification under constrained hardware conditions.
   - Integrated and optimized YNet, DLIB facial landmark tracking, and additional computer vision technologies for facial landmark analysis, eye tracking, head orientation estimation, gaze direction analysis, and attention/focus detection.
   - Developed hybrid AI pipelines combining neural-network inference with classical algorithmic computer vision approaches to achieve lower latency, improved stability, and reduced hardware requirements.
   - Customized and adapted existing ML models to project-specific requirements, including optimization for real-time inference speed, reduced CPU usage, minimized memory footprint, ROI-based processing, scene-aware adaptive inference, and improved accuracy for cybersecurity-oriented detection scenarios.
   - Participated in tuning, profiling, optimization, integration, validation, and production deployment of ML inference pipelines operating continuously in real-world enterprise environments.

Virtual Webcam Driver & Video Stream Infrastructure (Windows)
   - Developed Windows webcam virtualization driver and stream replication layer capable of splitting a single physical webcam stream into multiple synchronized virtual camera devices.
   - Solved Windows-exclusive webcam locking limitations where camera devices are often opened in exclusive-access mode, preventing simultaneous access by multiple applications.
   - Enabled several independent applications, AI modules, security analysis pipelines, and monitoring systems to consume the same live video stream simultaneously without hardware conflicts or additional physical devices.

Cloud Infrastructure / Enterprise Security / Backend Architecture
   - Participated in architecture and security design of enterprise-grade cloud-connected infrastructure combining autonomous workstation-level AI analysis with centralized real-time monitoring, policy management, security analytics, and organization-wide threat visibility.
   - Collaborated on the design of distributed cybersecurity architecture where AI-powered monitoring agents operated locally on protected workstations while securely communicating with centralized cloud infrastructure used by cybersecurity officers and security teams for live monitoring, behavioral analytics, incident response, and forensic analysis.
   - Participated in the design of enterprise security-policy framework enabling centralized configuration of AI detection rules, behavioral policies, privacy restrictions, risk thresholds, monitoring zones, escalation logic, and organization-wide cybersecurity enforcement strategies.
   - Designed and implemented highly scalable secure communication layer optimized for extremely high concurrency and low-latency telemetry delivery, capable of maintaining simultaneous communication with up to tens of thousands of connected clients while minimizing bandwidth and hardware utilization.
   - Developed custom high-performance secure communication protocols inspired by TLS/SSL concepts, utilizing AES encryption, public/private key cryptography, token-based authentication, secure session negotiation, integrity validation, and encrypted command/data channels optimized specifically for real-time AI telemetry and cybersecurity workloads.
   - Participated in the design of secure infrastructure for handling highly sensitive visual and behavioral data obtained from monitored environments, including protection of video streams, telemetry, snapshots, metadata, authentication tokens, and stored analytics in accordance with strict enterprise security requirements.
   - Contributed to security architecture decisions involving protected containerized environments, encrypted storage systems, secured service-to-service communication, isolated processing layers, and defensive mechanisms designed to reduce attack surface and prevent leakage of sensitive information.
   - Designed the platform to support both fully autonomous offline operation and cloud-connected enterprise deployment modes, allowing customers to operate securely even in isolated or bandwidth-restricted environments while preserving centralized visibility and policy synchronization capabilities when connectivity was available.


   🔧 Tools: C++ (Visual Studio / STL / WinAPI / multithreading), Python, OpenCV, DNN / Neural Networks, ONNX, YOLO, YNet, DLIB, DMA / zero-copy pipelines, OpenSSL / AES encryption, Docker, distributed enterprise systems, high-performance client-server architectures; OS: Windows (client-side AI engine, drivers, desktop services), Debian Linux (backend infrastructure, APIs, telemetry, cloud services, containerized environments).

🔗

🧑‍💼 Senior Software Engineer (AI and LLM focus)

🏢 1x1 IT services.

📅 Feb 2024 - May 2024 (3 months)

   - Developed an advanced system for generating detailed product descriptions using Large Language Models (LLMs), automating the creation of high-quality descriptions from product titles and models for a database of 100,000+ products.
   - Leveraged AWS Bedrock Cloud APIs alongside local GPT4All LLMs to balance performance, text quality, and cost-effectiveness, achieving optimal results with AWS’s Claude 3 Sonnet LLM while incorporating local GPT4All LLMs to manage expenses.
   - Employed Python as the primary programming language for integrating cloud and local LLM capabilities, implemented dynamically scaled network of crawlers/parsers.

   🔧 Tools: AWS Bedrock cloud API + GPT4All local LLMs, Python

🔗

🧑‍💼 Senior Software Engineer (C++, Linux, Big Data, AI)

🏢 Shopedia/Shopit AB, a Big-Data Company

📅 July 2022 - December 2023 (17 months)

Project description:
   Big data system to crawl the whole Internet, identify and catalog products.

Responsibilities:
   - Delivered significant speed improvement and robustness of a distributed Linux-based Big-Data application for analyzing, context-linking, and categorizing large scales of textual information stream acquired from web crawlers.
   - Implemented text context analysis, recognition of human-readable text in any language, and conversion to structured, tokenized data for AI processing.
   - Utilized CUDA for text tokenization by dictionary and other performance-critical tasks, enhancing processing speed.
   - Integrated OpenCV for image processing tasks, improving the system's ability to analyze visual data.
   - Integrated detectron2 and EfficientDet AI models to identify and classify objects in images, enhancing product cataloging capabilities.
   - Incorporated TrOCR for efficient text extraction from images, improving data collection from visual sources.
   - Worked with TensorFlow to develop and optimize neural network models for various classification and recognition tasks.

   🔧 Tools: GCC/CLang C/C++, QT Creator, SQL, Python, CUDA, OpenCV, OpenCL, TensorFlow, detectron2, EfficientDet, TrOCR, NoSQL, Atlassian Confluence, GIT/GitLab, JIRA, Debian OS.

🔗

🧑‍💼 Senior Software Engineer (C++, Python, PHP, Windows, Linux, Web, IoT)

🏢 Vladonai Software - Productivity Software

📅 June 2009 - now (17 years 1 month)

Project description:
   - [Project #1] AllMyNotes Organizer software title, localized to 20+ languages, with skinned UI and sound themes.
   - [Project #2] Smart Home system. Development and maintenance of various software titles, web presence, and IoT solutions.
   - [Project #3] Real-time trading bot for automated and AI-driven cryptocurrency exchange.

Responsibilities:
   - Drove the full life-cycle of the software title AllMyNotes Organizer - the implementation of my vision of how end-user software should look and work: smart, efficient, and easy to use. The program is full of interesting technical solutions. Localized to 20+ languages. Has totally custom, skinned UI, including very advanced rich text editor, with tables editing UI/logic functionality. Sound themes (beta version). A self-healing database engine, which is able to restore data even after very heavy damage of file logical structure (we have nearly zero cases of data corruption). Though it stores data locally, the program has strong back-end on our server, which, among other tasks, controls software activations and we have zero piracy.
   - The web-site www.vladonai.com (also allmynotes.org and allmynotes.info) are hosted on Debian Linux server. Server administration duties, based on Nginx/Apache web server, Exim email server, Exim, MySQL, WireGuard VPN, IPtables firewall, fail2ban intrusion-prevention system. Implemented online manual reader, RSS feed for list of updates and press-releases.
   - Created web-based control panel for the web-site, which displays very advanced analytics on program usage, site visitors, monitor web-site visitors in real-time (can see humans and bots), allows easily list all sales, edit user records, send email newsletters, submit PAD files to software catalogs automatically (a crawler bot for defined list of sites), monitor server resources and checks for major errors in real-time, monitor major site pages for errors, allows us quick generation of activation codes and vouchers, automatically schedule discounts/promos for various geolocations.
   - Created automated real-time trading software for cryptocurrency exchanges (Bittrex and Poloniex) for a proprietary client. The solution involved developing a fully custom neural network model based on TensorFlow algorithms, trained on proprietary historical trading data, capable of predicting short- and long-term exchange rate changes in real time.
   - Development of a Smart Home system based on a mesh network of IoT micro-controllers to automate work with automatic surveillance cameras (OpenCL), efficient house energy flow distribution from Solar Station + Grid to smart consumption, including V2H (Vehicle to House) mode for Electric cars.

   🔧 Tools: Visual Studio C++, GCC, PHP/Python/Perl/Node.js/JavaScript, SQL, Git/Gitea, Apache, Nginx, ESP32/VxWorks/Raspberry Pi, Mantis Bug Tracker, TensorFlow, Windows, Debian OS, Docker, RTOS, Embedded Linux.

🔗

🧑‍💼 Software Developer (C++, Mobile, Windows)

🏢 Iambic Inc. - Mobile Software - Sunnyvale, CA

📅 March 2001 - May 2009 (8+ years)

Project description:
   Mobile Software.

Responsibilities:
   - Worked on the TinySheet project, an Excel-like application for Palm OS. Added charting support, made it Excel-compatible with seamless synchronization of Excel files - we were the only company among the competition able to 'natively' synchronize charts between Excel and Palm. Built a UI library to support the new high-resolution devices (no easy task at the time - literally we had to write 'drivers' for every possible device on the market).
   - Developed a so-called 'Intellectual Text Processor' - a text-parsing library/API able to automatically recognize and identify information of various types out of plain text: address line, country, city, ZIP, first name, last name, company name, job title, phone numbers, e-mails, URLs, IM handles, etc. It performed batch text recognition on business cards, email signatures, and other irregularly formatted texts. Ported it to multiple platforms.
   - Raised to position of leading developer of the company, worked on all products, including Agendus for Palm (for Palm, and Synchronization Conduit), TripTracker (version 4 for Windows and WM), Vehicle Manager (for Win and WM), Health and Diet (for Win and WM), Iambic Mail (for Palm, and Synchronization Conduit for Palm), TinySheet synchronization conduit, FastWriter (for Palm and synchronization conduit), Iambic Reader (for Palm and synchronization conduit), Time Reporter (for Palm and synchronization conduit), Sales Warrior (for Palm). All these projects were supported mostly by me when was need to add new feature or fix some bug, lead support for other team members, did code reviews, also I was dealing directly with QA and support staff in order to solve problems related to these products.
   - I was engaged in bundling TinySheet (our Excel-like application for Palm OS), FastWriter (our Word-like application for Palm OS), and the Iambic Mail application into a single 'Iambic Office' package on the PC side. I built the shared synchronization manager application and improved the conduits (synchronization modules) so all these applications worked together seamlessly as one package.
   - Implemented plugins for Outlook, Excel and Word for quick access to Synchronization with Mobile functions. It's done using Visual Basic and C++ API of Outlook
   Note: for first 3 years I was working in French branch office of Iambic Inc, and after that office was closed moved to country of my origin, Ukraine, and continued working for Iambic, Inc. as independent contractor, remotely, visiting Iambic Inc headquarter office in California approximately once a year. I was a leading developer for most of the software titles developed by Iambic, Inc - I was working for Iambic, Inc. for 7+ years, which is an absolute record for this company. Worked there up until the last day of the company, when it was sold to our competitors... The company was successful and profitable for the period when I was working there, I was loyal to it until the end.

   🔧 Tools: Code Warrior C++, Visual C++, Visual Basic, MFC, Palm OS API, and Win32 API.

🔗

🧑‍💼 Software Developer (C++, Mobile)

🏢 Yellow Computing - Mobile Software

📅 May 2000 - March 2001 (10 months)

Project description:
   Mobile Software.

Responsibilities:
   - Created dictionary application mobile software title Pro'Lingua for Palm Computing Platform version 2 (created it from ground up, not based on v1), it allowed quick translation of words between various language pairs. The biggest technical challenge in this project was developing a new compression algorithm to compress dictionaries on the mobile device. It has a high compression ratio and a low memory footprint. This project is also completely object-oriented (even message handling and all UI object wrappers), which was a very rare case for the Palm platform.

   🔧 Tools : Code Warrior C++, GCC, Delphi/Pascal, Palm OS API, and Win32 API.

🔗

🧑‍💼 Software Developer (C++, Mobile, Windows)

🏢 MicroLogic Software Corp (USA, remote) - Hackensack, NJ

📅 December 1998 - September 2000 (2 years)

   - Development of the CAD program 'Machine Shop' (www.emachineshop.com). This CAD application is used to draw technical drawings and send it via the Internet to MicroLogic for manufacturing. Developed a large part of the user interface, the client/server API for placing orders with CAD drawings, server-side order storage, a secure/encrypted communication channel using the PGP algorithm, export of data to .DXF (AutoCAD) format, and printed output of technical drawings.
   - Developed a multi-platform algorithm to synchronize hierarchical trees (N-way synchronization) - PC to Mobile and Mobile to CP, also between multiple PCs within LAN network. Which was used in number of product of the company.

   🔧 Tools: Code Warrior C++, GCC, Visual C++, Borland C++, MFC, Palm OS API, and Win32 API.

🔗

🧑‍💼 Various/Freelance Application Developer

🏢 Self

📅 July 1997 - December 1998 (18 months), and occasionally in the 2000s between jobs

   - Epoch Clock application to extend Windows Clock in the system tray, added calendar/reminders capability to it.
   - BeCrossword PalmPilot and WindowsMobile game to resolve puzzles on the go.
   - Simple CAD application to draw geographical maps. Allows drawing and viewing of maps, add/edit/remove map elements (houses, roads, parks, pipes, etc.), filter specified elements, search, scaling and some more. This project was done for Ternopil state water supply company to keep information about their objects on the map.
   - Phone Book with list of all phone numbers in Ternopil city. This application has nice looking user interface (my first attempt to create something not ordinal, and quite successful), can add, delete and modify data in database, has powerful search (search in any field, also possible to use wildcards), sorting data by different fields, has very secure data storing format (in case of partial database damage in most of cases can be restored).
   - Warehouse accounting automation. Used to automate all warehouse accounting and other operations in private firm 'Konist' (which is probably #1 regional company which is selling electrical equipment such as pumps, engines, etc.), application includes relational database support for different needs, search, filtering, statistics, printing and some more.
   - Drugstore accounting automation. Used to automate most of drugstore accounting operations in Ternopil drugstore #160, includes relative databases, search, filtering, statistics, printing and some more.
   - Fractal explorer app. Allow to explore fractals, users can set fractal parameters, scale, see statistics and more.
   - Developed local companies catalogue CD by order of Infotechcentre company in 1998.
   - Developed first online shop web-site for Infotechcentre company in 1998. That site was fully developed, it was a supported CMS system long before commercial CMS systems appeared on a market, it was up for 15 years until company was in business.
   - Created the first official website for Ternopil State Technical University, which remained the university's online presence for 5 years.

   🔧 GCC, Visual C++, MFC, Win32 API, HTML/JavaScript, Delphi/Pascal.

🔗

🏢 Ternopil State Technical University

📅 1997-2002 (5 years)

   Degree: Specialist of Computers Sciences (equivalent of Master degree, 5 years university study). Learning Data Structures and Algorithms, Mechatronics/Robotics, Automation Engineering, Math Analysis. Also, I've started working on a second year of university, studied to apply learned skills on practice.

🔗

🧑‍💼 Self-study

📅 1991 - 1997

   I've been programming since I was 11 years old. I genuinely love it and spend a lot of my time on it - I practically live on the Internet. I have a deep understanding of how to write code the right way: all my code is written in great style, and things like naming, security and clarity matter a great deal to me. My first computer was a Sinclair ZX-Spectrum that I built largely with my own hands - many parts were soldered manually - and thanks to that I have a solid understanding of computer hardware. This background lets me understand software crystal clear, as a whole process, because I know the structure of the processor, the computer's parts, and how they interact with the CPU and each other at a very low level (I know exactly how RAM is addressed on the address bus and read over the data bus, the difference between memory and I/O addressing; my first programs were saved on magnetic tape and displayed on an old-fashioned CRT TV).

🔗   I'm an old-school computer/C++ geek :-) I spend most of my time in front of my PC - administering my Linux server and its services, tinkering with and developing the IoT system in my house, constantly sharpening my development skills and looking for new, exciting ways to solve problems faster and more efficiently. I love art, music, spiritual growth, occasional traveling, and - once again - computers :) Even when I'm not working, I find myself thinking about a problem I'm working on...

   I've developed, entirely on my own, a complete code library that makes GUI development dramatically faster. It has full skinning support (WinAmp-like, but much stronger) and effortless localization (users can even localize the program into their own language and share language modules without recompiling anything - it can be done in plain Notepad by an end user), and its security and development speed are far superior to traditional approaches. Across all software development I put most of my attention on the user interface, because the eye-catching factor is extremely important for a product's success - and this library is the result of my best experience. You can see it in action in the AllMyNotes Organizer product.

   Maintaining a healthy and substance/alcohol-free lifestyle, emphasizing focus, productivity, and a commitment to well-being.

🔗
    github.com/Frytskyy

🔗

📅 2025–2026

Project description:
   Terminal server health dashboard - single Python file, no Docker, no Grafana.

Responsibilities:
   Multi-threaded ping + HTTP monitoring, adaptive polling, hardware-bound credential encryption (Fernet/PBKDF2), Telegram/SMTP/Webhook alerts, braille sparkline graphs, psutil system metrics.

   🔧 Python, psutil, Fernet, PBKDF2, Telegram Bot API  ·  Cross-platform: Windows, Linux, macOS

🔗

📅 2025

Project description:
   VPN configurator for star topology networks.

Responsibilities:
   Cryptographic key management, IP conflict detection, QR code generation for mobile clients, config encryption (Fernet), network cloning/templates.

   🔧 Python, WireGuard, Fernet, qrcode

🔗

📅 2025

Project description:
   Git-based multi-device sync with automatic Gitea repo creation and project metadata storage.

Responsibilities:
   Guided setup wizard, client + server pair architecture.

   🔧 Python, Git, Gitea API, Docker, PostgreSQL

🔗

📅 2024

Project description:
   Interactive terminal menu for daily Git routines.

Responsibilities:
   Commit/push, branch management, checkout, tags, remote config. subprocess orchestration, JSON config persistence, colorama UI.

   🔧 Python, subprocess, colorama, JSON

🔗

📅 2010-now (work in progress - coming soon for beta testing)

Project description:
   Unified terminal control plane for multi-server infrastructure monitoring, cybersecurity automation, and LLM-powered decision-making.

Responsibilities:
   Architecture: Single-file, agentless monolithic TUI application (40,000+ LOC Python) managing unlimited heterogeneous Linux infrastructure exclusively via paramiko SSH-no agents required on servers. Async-native coroutine pool orchestrates real-time data collection, command execution, and event processing across server fleets. Hardware-bound encrypted credential vault (Fernet + PBKDF2) ensures portable, secure key management.

Core Capabilities: Real-time infrastructure monitoring (CPU, RAM, disk, network, Docker, databases), active attack detection and firewall management (UFW/iptables), service lifecycle control, cross-host log aggregation with regex search, SFTP file manager, deployment orchestration (git pull, restart, rollback), backup/restic status tracking, cron/systemd timer management, MySQL/PostgreSQL/Redis slow-query analysis, email monitor, and process lifecycle audit (kill, OOM detection, CPU throttling).

AI-Enhanced Autonomous Operations: Unified AIEngine middleware (Claude, GPT-4, Gemini, Mistral, Groq, LM Studio support) integrates with every operational screen. Per-host LLM instructions enable background loops for autonomous governance (e.g., "restart MySQL if slow-query rate exceeds 10/min"). Natural-language command terminal (Ctrl+A) translates English descriptions to production CLI/SSH/Bash. Threat assessment analysis contextualizes attack patterns; health audit performs cross-host root-cause analysis; defense mode auto-executes mitigations (bans, restarts) within policy limits. F8/F9 shell integration provides inline AI hints and command translation. Dual-tier model strategy (SMART for reasoning, FAST for classification) optimizes cost and latency.

Security & DevOps Excellence: Attack radar visualizes malicious IPs, geo-location, and timeline; 25+ screens cover every infrastructure dimension; bulk operations enable scripted fleet changes; network reconnaissance presets; API-rate-aware token budgeting; and comprehensive security wizard onboarding. All SSH credentials encrypted at rest; flexible config (JSON-based); portable export/import capability.

Technical Distinction: Purpose-built for infrastructure engineers who demand real-time visibility and autonomous defense without sacrificing control-combining DevOps observability, cybersecurity response, and emerging LLM-powered decision support into a single, cohesive, agentless platform.

   🔧 Python 3.10+, Textual TUI, asyncio, paramiko, cryptography (Fernet/PBKDF2), Claude API, OpenAI/Gemini/Mistral, SSH, Linux kernel

🔗

📅 work in progress - coming soon for beta testing

Project description:
   Interactive terminal firewall management tool for Linux servers.

Responsibilities:
   Fortress mode (deny-all-except-specified), manual rule creation with syntax validation, rule editing/deletion by chain, safety checks, enhanced comment system with JSON storage, config import and security analysis.

   🔧 Python, Rich TUI, threading, iptables, iptables-persistent

🔗

📅 work in progress - coming soon for beta testing

Project description:
   Windows system tray application combining a cron-like task scheduler with real-time security monitoring.

Responsibilities:
   Detects keyloggers, clipboard hijacking, screen capture threats, and suspicious autorun entries. Features startup task management, periodic job execution, password-protected tray menu, and comprehensive event logging with 90-day retention.

   🔧 Python, pystray, Windows API, threading

🔗

📅 work in progress - coming soon for beta testing

Project description:
   Secure SSH connection manager with encrypted credential storage.

Responsibilities:
   Hardware-bound encryption ties stored passwords to the local machine via disk serial fingerprinting (PBKDF2-HMAC-SHA256 → Fernet), making leaked config files useless on any other hardware.

   🔧 Python, Fernet, PBKDF2-HMAC-SHA256  ·  Cross-platform: Windows, Linux

🔗
    View my public GitHub profile - it includes a number of open-source DevOps projects as well as C++ code samples.