Full Stack Software Engineer
📧 Email: serg.yalosovetsky@gmail.com
📱 Telegram: @ozzzimandiya
🌐 GitHub: https://github.com/serg-yalosovetsky
🔗 LinkedIn: https://www.linkedin.com/in/serhii-yalosovetskyi/
📃 CV: pdf
Summary
Full Stack Software Engineer specializing in Python, AI/LLM systems, Flutter, and Svelte. Currently building production-grade AI applications featuring local LLM inference, RAG pipelines, and semantic search over real-world data. Developed scalable backend systems and mobile applications, leveraging DevOps practices that improved deployment speed by 40%, leading to quicker product iterations based on user feedback in agile environments.
-purple){kind=icon}
Technical Skills
🤖 AI & LLM Systems
🛠 Back-End
📱 Mobile Development
💻 Front-End
⚙️ Tools & Platforms
🧪 Testing & Frameworks
🔬 Research & Work Experience
Senior Full Stack Python + DevOps
Hermes Logistics Technologies
April 2025 – November, 2025
- Spearheaded the complete migration of the entire backend infrastructure from legacy Magic XPA systems to FastAPI, resulting in a modern, scalable, and high-performance architecture.
- Architected and implemented comprehensive CI/CD pipelines in Azure DevOps, fully automating testing and deployment processes to streamline release cycles.
- Designed a secure and scalable deployment infrastructure on AWS ECS, incorporating automated rollback strategies to ensure failproof releases and high availability.
Software Developer
Techwhirl Infotech
October 2024 – January 23, 2025
- Engineered a robust backend for a Telegram-based crypto exchange using FastAPI and Piccolo, reducing transaction times to under 2 seconds while maintaining adherence to evolving blockchain regulations.
- Collaborated with cross-functional teams to define precise business requirements, enabling the rapid achievement of key milestones in just two months despite evolving project parameters.
Software Developer
Zeus Electronics
2021 – October 2024
- Redesigned architectural frameworks within Fyuzd, reducing monthly downtime incidents from six to one and ensuring smooth integration with existing workflows.
- Transformed legacy systems into scalable, cloud-native architectures while aligning with industry standards, directly supporting the company’s strategic objectives.
- Executed comprehensive performance and reliability tests on REST APIs, pinpointing bottlenecks and driving a 50% acceleration in response times across endpoints.
- Configured Kubernetes environments to streamline Allgram CI/CD pipelines.
- Analyzed requirements across three applications and collaborated with stakeholders to refine specifications, achieving a 25% reduction in initial development time.
Cybersecurity Specialist
Vodafone Ukraine
2018 – 2021
- Performed in-depth API security assessments by performing over 30 simulated attacks to evaluate system robustness and inform new guidelines aligned with OWASP standards.
- Established security protocols based on OWASP Mobile Security and Web App Security Guides, reducing critical incident response times by 40%.
- Fortified the organization’s cyber defense strategy by implementing tools such as McAfee Endpoint Security and VMware ONE UEM, protecting sensitive data for over 5,000 daily users.
- Implemented advanced scanning methods to identify structural flaws, resolving key issues that impacted service delivery for thousands of daily interactions.
- Created a custom captcha system to neutralize DDOS attacks, ensuring uninterrupted service for more than 50,000 daily mobile interactions.
- Automated daily workflows with Bash and VBA scripting, saving over 60 work hours monthly.
- Spearheaded the formulation of business requirements for five major cybersecurity projects, ensuring alignment with strategic goals and enhancing cross-department collaboration.
🌱 Projects
✅ Summagram (2025 – present)
AI assistant for personal Telegram history, built end-to-end as a production system:
- Designed an ETL pipeline that ingests full Telegram chat history into PostgreSQL with incremental sync and real-time SSE progress streaming.
- Built a hybrid RAG retrieval engine combining semantic search (ChromaDB embeddings) with lexical full-text search and NLP-based query expansion.
- Integrated local LLM inference via vLLM and SGLang (text + vision models) with automatic GPU/CPU fallback and dynamic model orchestration.
- Implemented an NLP entity extraction pipeline (persons, topics, dates) enabling intelligent topic-based chat routing when no explicit contact is named.
- Stack: Python, FastAPI, PostgreSQL, ChromaDB, vLLM, SGLang, Docker, Next.js.
United Help Ukraine
✅ UnitedHelp
- Established the infrastructure for both web and mobile applications designed to enhance volunteer coordination, with real-time tracking that accelerated task completion.
- Crafted innovative backend features to address key challenges at UnitedHelp, prioritizing tasks to resolve three critical issues affecting user satisfaction.
- Designed and launched a user-friendly Flutter mobile platform, streamlining assignment tracking and reducing coordination time by one-third.
✅ Captcha Project
- Developed a Python tool for generating and verifying captchas, bolstering web application security.
✅ SongLyrics App
- Engineered features in a Flutter app that allow users to search and store song lyrics offline, resolving three major performance challenges based on user feedback.
✅ Portfolio Website
- Built a SvelteKit-powered site featuring a landing page, portfolio, about section, request forms (with Telegram integration), and a blog.
🎓 Education
Bachelor’s Degree in Informatics and Computer Engineering
National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”
2014 – 2018
Language Skills
- English: B2 (Upper-intermediate).
- Ukrainian: Native.
📞 Contact
📧 serg.yalosovetsky@gmail.com
📱 @ozzzimandiya
🌐 GitHub Profile